AperTO - Archivio Istituzionale Open Access dell ...

32
14 January 2022 AperTO - Archivio Istituzionale Open Access dell'Università di Torino Original Citation: The Neural Correlates of Time: A Meta-analysis of Neuroimaging Studies Published version: DOI:10.1162/jocn_a_01459 Terms of use: Open Access (Article begins on next page) Anyone can freely access the full text of works made available as "Open Access". Works made available under a Creative Commons license can be used according to the terms and conditions of said license. Use of all other works requires consent of the right holder (author or publisher) if not exempted from copyright protection by the applicable law. Availability: This is a pre print version of the following article: This version is available http://hdl.handle.net/2318/1723185 since 2020-01-15T15:24:42Z

Transcript of AperTO - Archivio Istituzionale Open Access dell ...

Page 1: AperTO - Archivio Istituzionale Open Access dell ...

14 January 2022

AperTO - Archivio Istituzionale Open Access dell'Università di Torino

Original Citation:

The Neural Correlates of Time: A Meta-analysis of Neuroimaging Studies

Published version:

DOI:10.1162/jocn_a_01459

Terms of use:

Open Access

(Article begins on next page)

Anyone can freely access the full text of works made available as "Open Access". Works made available under aCreative Commons license can be used according to the terms and conditions of said license. Use of all other worksrequires consent of the right holder (author or publisher) if not exempted from copyright protection by the applicable law.

Availability:

This is a pre print version of the following article:

This version is available http://hdl.handle.net/2318/1723185 since 2020-01-15T15:24:42Z

Page 2: AperTO - Archivio Istituzionale Open Access dell ...

The Neural Correlates of Time: A Meta-analysisof Neuroimaging Studies

Andrea Nani1,2, Jordi Manuello1,2, Donato Liloia1,2, Sergio Duca1,2,Tommaso Costa1,2, and Franco Cauda1,2

Abstract

■ During the last two decades, our inner sense of time hasbeen repeatedly studied with the help of neuroimaging tech-niques. These investigations have suggested the specific in-volvement of different brain areas in temporal processing. Atleast two distinct neural systems are likely to play a role in mea-suring time: One is mainly constituted by subcortical structuresand is supposed to be more related to the estimation of timeintervals below the 1-sec range (subsecond timing tasks), andthe other is mainly constituted by cortical areas and is supposedto be more related to the estimation of time intervals above the1-sec range (suprasecond timing tasks). Tasks can then be per-formed in motor or nonmotor (perceptual) conditions, thusproviding four different categories of time processing. Our

meta-analytical investigation partly confirms the findings of pre-vious meta-analytical works. Both sub- and suprasecond tasksrecruit cortical and subcortical areas, but subcortical areas aremore intensely activated in subsecond tasks than in supra-second tasks, which instead receive more contributions fromcortical activations. All the conditions, however, show strong ac-tivations in the SMA, whose rostral and caudal parts have an im-portant role not only in the discrimination of different timeintervals but also in relation to the nature of the task conditions.This area, along with the striatum (especially the putamen) andthe claustrum, is supposed to be an essential node in the dif-ferent networks engaged when the brain creates our sense oftime. ■

INTRODUCTION

The passage of time is one of the most pervasive aspectsof our lives. Intuitively, everyone feels to know what itmeans that time passes, but if we were asked to expressin words the nature of time and the feeling of temporal-ity, we would be in the same situation as Saint Augustine,who declared in his Confessions: “What then is time? Ifno one asks me, I know what it is. If I wish to explainit to him who asks, I do not know.” Saint Augustine wrotebetween the IV and V centuries AD, but to date the na-ture of time remains elusive. Time has been consideredan essential feature of reality, on the one hand (Unger &Smolin, 2015), or a figment of the way we know reality,on the other (Rovelli, 2018). Independent of the fact thattime is a real constituent of the outside world or just afabrication of our inner world, the brain is unquestion-ably involved in perceiving or constructing the temporalframework within which events (seem to) happen andinteract.During the last two decades, time perception (or

construction) has been repeatedly investigated with thehelp of neuroimaging techniques. These studies haveput forward different results, which in turn have beenused to support different theories about the involvement

of specific brain areas in temporal processing. These theo-ries can be divided into two types: dedicated and intrinsicmodels (Ivry & Schlerf, 2008). The dedicated models hy-pothesize a specific mechanism exclusively devoted to theestimation of time; in contrast, the intrinsic models assumethat various brain structures normally associated with sen-sory and cognitive processes can also act as interval timers(Wittmann, 2013). A classic example of a dedicated mech-anism for time perception is the pacemaker-accumulatorsystem (Taatgen, van Rijn, & Anderson, 2007; Zakay &Block, 1997). The idea is that the neural pacemaker gener-ates a series of pulses that, just like the ticks of a clock, arerecorded to measure time intervals. In particular, the dopa-minergic system and the BG, especially the striatum (cau-date nucleus and putamen) and the substantia nigra parscompacta, are thought to be the underpinnings of thisneural pacemaker (Meck, Penney, & Pouthas, 2008;Buhusi & Meck, 2005). In particular, the striatum, which re-ceives inputs from a large number of cortical neurons,would be the main region involved in detecting the oscilla-tory patterns or frequencies related to the activity of differ-ent cortical sites. Neuroimaging investigations that applieddifferent techniques (fMRI, PET) have supported the pivotalrole played by the striatum in time processing (Harrington,Zimbelman, Hinton, & Rao, 2010; Coull & Nobre, 2008;Rao, Mayer, & Harrington, 2001). However, other experi-ments provide evidence that, according to different tasks1GCS-fMRI, Koelliker Hospital, Turin, Italy, 2University of Turin

© 2019 Massachusetts Institute of Technology Journal of Cognitive Neuroscience X:Y, pp. 1–31https://doi.org/10.1162/jocn_a_01459

Page 3: AperTO - Archivio Istituzionale Open Access dell ...

and conditions, other several brain areas are supposed tobe involved in the perception of time.

Various mechanisms underlying intrinsic models havetherefore been proposed. The most general mechanismassumes that short-term synaptic plasticity, which is atthe basis of the dynamics of neural states, can transformthe brain networks in time keepers (Buonomano &Maass, 2009). According to this view, every cerebral areahas the capacity to process and estimate time. Other pro-posals, on the contrary, are more specific and pinpoint acertain brain region or neural population as the substrateof a particular temporal task. For instance, the right pFC(Lewis & Miall, 2006a, 2006b), the cerebellum (Ivry,Spencer, Zelaznik, & Diedrichsen, 2002), the SMA (Macar,Coull, & Vidal, 2006), the right inferior parietal lobule(Battelli, Pascual-Leone, & Cavanagh, 2007), and the anteriorinsula (Craig, 2009a) have been variously associated withtime perception, judgment, and the processing of intervals’duration.

The variegated picture of our inner sense of time cre-ated by the involvement of these various brain structuresis probably due to different experimental contexts andsettings, the characteristics of the tasks employed, theduration of stimuli, the nature of responses, as well asthe application of different statistical procedures. For ex-ample, Lewis and Miall (2003a) carried out a label-basedreview of neuroimaging studies on time perception. Theauthors divided the experimental results according to theinterval duration (i.e., subsecond or suprasecond tasks),the use of movement (i.e., motor or nonmotor tasks) andthe nature of response (i.e., automatic or cognitively con-trolled tasks).

With regard to motor timing research, the task that ismost frequently used is paced finger tapping: Participantsare asked to tap a key following the rhythm of an isochro-nous metronome and then continue tapping at the samepace in absence of stimulation. In some experiments, thecontinuation phase is not required, but the participantsare asked to tap at the middle point between groups ofpacing stimuli. Other nonmotor experiments use the in-terval reproduction task, in which participants recall theduration of an interval between two stimuli to which theywere previously exposed. A variation of this task is to askthe participants directly, without exposure, to produce aspecific interval (i.e., “produce an interval of 6 sec”). On theother hand, time perception tasks require participants tojudge the span of intervals. Frequently, participants areasked to compare and discriminate the duration of two suc-cessive stimuli by judging which of them is longer than theother. In these experiments, the duration of the stimuli tobe judged is very variable, ranging from 200 to 24 sec.According to these different timing tasks, different brainstructures are recruited (Buhusi & Meck, 2005; Mauk &Buonomano, 2004; Lewis & Miall, 2003a). Furthermore,neuropsychological studies in humans and research on an-imals point out that the estimation of temporal intervalsabove and below 1 sec relies on the activation of different

neural mechanisms and assemblies (Harrington & Haaland,1999; Gibbon, Malapani, Dale, & Gallistel, 1997).Overall, with regard to these different experimental

conditions, studies have suggested that at least two dis-tinct neural systems are likely to play a role in measuringtime: One is mainly constituted by subcortical structuresand is supposed to be more related to subsecond timingtasks, and the other is mainly constituted by cortical areasand is supposed to be more related to suprasecond tim-ing tasks.The distinction between different tasks and time inter-

vals has been also highlighted by the meta-analysis ofWiener, Turkeltaub, and Coslett (2010), which analyzedfour different conditions: subsecond motor, subsecondperceptual (i.e., nonmotor), suprasecond motor, and su-prasecond perceptual. We have decided to adhere to thisdistinction, as it describes better the current literature ontime perception studies. It should be observed, however,that it has some limitations, especially with regard to thecategory of the “perceptual” condition. In fact, rigorouslyspeaking, this category is not strictly perceptual, as its ex-perimental tasks include also the engagement of higherorder cognitive functions, such as attention and salience.As Wiener et al. (2010), we use therefore “perceptual” ina broad sense as a synonym of “nonmotor,” so as to des-ignate all the experiments in which the task requirementis not strictly motor. For the studies that tested an inter-val across subsecond and suprasecond stimuli, we alsodecided, like Wiener et al. (2010), to determine whethermost of the interval was above or below 1 sec and attri-bute them to the corresponding category. Furthermore,with regard to these possible confounding factors, we ap-plied the fail-safe techinque to assess the robustness ofour results (see Methods for a detailed explanation of thisprocedure).The results of the meta-analysis by Wiener et al. (2010)

supported the idea that certain neural structures are in-volved in subsecond timing tasks and others in supra-second timing tasks. In particular, the analysis ofsubsecond tasks revealed a greater activation of subcorti-cal sites, such as the bilateral cerebellum and left BG. Incontrast, the analysis of suprasecond tasks principallyshowed the activations of cortical areas, such as the bilat-eral SMA, frontal cortex, and cingulate gyri. Similarly, themeta-analysis carried out by Ortuño, Guillen-Grima,Lopez-Garcia, Gomez, and Pla (2011) largely reproducedthese findings. However, in this case statistical analyseswere performed on independent studies and only consid-ered the cerebral activations related to implicit or explicittime estimation tasks. Finally, Schwartze, Rothermich, andKotz (2012) focused their meta-analysis of fMRI studies ona subarea of the SMA by contrasting different temporalprocessing functions, such as sensory, sensorimotor, se-quential, nonsequential, explicit, nonexplicit, subsecond,or suprasecond tasks. These authors identified differentactivations of the SMA, which might be correlated with dif-ferent types of temporal processing.

2 Journal of Cognitive Neuroscience Volume X, Number Y

Page 4: AperTO - Archivio Istituzionale Open Access dell ...

Table 1. Articles Included in the Meta-analysis Using the PRISMA Statement International Guidelines

Study Tool Participants FociExperimental Task

(Contrast or Condition) Duration (msec)

Aso et al. (2010) (Exp. A) fMRI 14 6 Temporal reproduction (maineffect of motor timing)

500–999

Bengtsson et al. (2005) fMRI 7 6 Paced finger tapping andarticulation (more vs. lesscomplex)

375–750–1125

Bueti, Walsh, et al.(2008) (Exp. A)

fMRI 14 11 Temporal reproduction(action vs. control)

300–600–900–1200

Berns et al. (1997) PET 10 3 RT sequence (more vs.less complex)

700

Jahanshahi et al.(2006) (Exp. A)

PET 8 11 Temporal reproduction(short vs. long intervals)

500

Jancke et al. (2000) fMRI 8 3 Paced finger tapping(paced vs. cued finger)

400

Jantzen et al. (2004) fMRI 14 10 Paced finger tapping (syncopationvs. synchronization)

800

Jantzen et al. (2005) fMRI 12 7 Paced finger tapping (syncopationvs. synchronization)

800

Jantzen et al. (2007) fMRI 9 7 Paced finger tapping (paced vs.cued finger)

800

Karabanov et al. (2009) fMRI 16 43 Temporal sequence performance(visual vs. auditory rhythms)

375–750–1125–1500

Lewis et al. (2004) fMRI 10 42 Paced finger tapping (paced vs.cued finger)

700

Lutz et al. (2000) fMRI 14 1 Paced finger tapping (isochronousvs. nonisochronous)

667

Mayville et al. (2002) fMRI 9 20 Paced finger tapping (syncopationvs. synchronization)

800

Murai & Yotsumoto(2016) (Exp. A)

fMRI 21 7 Temporal duration reproduction(sub vs. suprasecond)

400

Ortuño et al. (2002) PET 10 9 Temporal articulation (mentally vs.auditory rhythm counting)

1000

Oullier et al. (2005)(Exp. A)

fMRI 15 19 Paced finger tapping (syncopationvs. synchronization)

800

Oullier et al. (2005)(Exp. B)

fMRI 15 12 Paced finger tapping (synchronizationvs. syncopation)

800

Pastor et al. (2004) fMRI 14 6 Temporal discrimination(clicks vs. location)

5–140

Penhune et al.(1998) (Exp. A)

PET 11 5 Paced finger tapping (novelvs. learned)

250

Penhune et al.(1998) (Exp. B)

PET 11 4 Paced finger tapping (novelvs. learned)

750

Pope et al. (2005) fMRI 13 21 Temporal sequences reproduction(main effect of alternating force)

400–800

Rao et al. (1997)(Exp. A)

fMRI 13 7 Temporal discrimination(temporal continuation vs. pitch)

300

Nani et al. 3

Page 5: AperTO - Archivio Istituzionale Open Access dell ...

Table 1. (continued )

Study Tool Participants FociExperimental Task

(Contrast or Condition) Duration (msec)

Rao et al. (1997) (Exp. B) fMRI 13 7 Temporal discrimination(temporal continuation vs. pitch)

600

Schubotz & von Cramon(2001) (Exp. A)

fMRI 12 20 Temporal reproduction (temporalvs. control pressing)

290–1450

Subsecond motor timing condition: 21 articles, 24 experiments, 293 participants, and 287 foci of activation

Aso et al. (2010)(Exp. B)

fMRI 14 4 Temporal discrimination(main effect for sensory timing)

500–999

Bengtsson et al.(2009) (Exp. A)

fMRI 17 12 Auditory rhythm perception(rhythm vs. random)

200–400–600–800

Bengtsson et al.(2009) (Exp. B)

fMRI 17 7 Auditory rhythm perception(metric vs. isochronous)

200–300–700–900

Beudel et al.(2009) (Exp. A)

fMRI 18 10 Temporal estimation (timeahead vs. stop)

100–800

Beudel et al.(2009) (Exp. B)

fMRI 18 9 Temporal estimation (timeahead vs. spatial place)

100–800

Bueti, Walsh, et al.(2008) (Exp. B)

fMRI 14 4 Temporal estimation (perceptionvs. control)

300–600–900–1200

Coull & Nobre(1998) (Exp. A)

PET 15 4 Temporal perception (whenvs. where)

0–315

Coull & Nobre(1998) (Exp. B)

fMRI 15 3 Temporal perception (whenvs. where)

0–315

Ferrandez et al.(2003)

fMRI 11 12 Temporal discrimination (estimationvs. intensity of duration)

700

Gutyrchik et al.(2010)

fMRI 13 2 Temporal perception (durationvs. color)

100–250–1000–1300

Lewis & Miall(2003b) (Exp. A)

fMRI 8 20 Temporal discrimination(temporal vs. length)

600

Maquet et al. (1996) PET 9 6 Temporal discrimination(time vs. intensity duration)

700

Mathiak et al. (2004) fMRI 12 1 Temporal auditory discrimination(auditory tones duration)

200

Neufang et al. (2008) fMRI 34 1 Temporal discrimination (time vs.visual compatibility)

250–350

Schubotz et al. (2000) fMRI 20 14 Temporal discrimination (temporalvs. color pitch)

900

Schubotz & von Cramon(2001) (Exp. B)

fMRI 12 20 Temporal monitoring (temporal vs.control monitoring)

290–1450

Shih et al. (2009) fMRI 17 5 Temporal discrimination(temporal vs. control)

100–450

Shih et al. (2010) fMRI 21 13 Temporal discrimination(temporal vs. control)

400–450

Tipples et al. (2013) fMRI 17 8 Temporal estimation (duration vs.gender of stimuli)

400–700–1000–1300–1600

Tomasi et al. (2015) fMRI 36 27 Temporal prediction (temporal vs. 600

4 Journal of Cognitive Neuroscience Volume X, Number Y

Page 6: AperTO - Archivio Istituzionale Open Access dell ...

Table 1. (continued )

Study Tool Participants FociExperimental Task

(Contrast or Condition) Duration (msec)

memory attention)

Tregellas et al. (2006) fMRI 20 10 Temporal discrimination(difficult vs. easy)

200

van Wassenhove et al.(2011) (Exp. A)

fMRI 15 6 Temporal dilation estimation(looming vs. standard stimuli)

500

van Wassenhove et al.(2011) (Exp. B)

fMRI 15 8 Temporal dilation estimation(receding vs. standard stimuli)

500

Wittmann, van Wassenhove,et al. (2010) (Exp. A)

fMRI 20 6 Temporal dilation estimation(looming vs. standard stimuli)

500

Wittmann, van Wassenhove,et al. (2010) (Exp. B)

fMRI 20 8 Temporal dilation estimation(receding vs. standard stimuli)

500

Subsecond perceptual (nonmotor) timing condition: 20 articles, 25 experiments, 428 participants, and 220 foci of activation

Ackermann et al. (2001) fMRI 8 9 Paced finger tapping (isochronous trainsof clicks at different frequencies)

6000

Basso et al. (2003) fMRI 5 8 Temporal production (time vs.working memory control)

1500

Botzung et al. (2008)(Exp. A)

fMRI 10 11 Events evocation (past vs. semanticmemory condition)

3700–11240

Botzung et al. (2008)(Exp. B)

fMRI 10 12 Events evocation (future vs. semanticmemory condition)

3700–11240

Brunia et al. (2000) PET 8 4 Temporal production (correct vs.incorrect feedback)

3000

Coull et al. (2013)(Exp. A)

fMRI 27 22 Temporal production (temporal vs.neutral reproduction)

600–1000–1400

Dudukovic & Wagner(2007)

fMRI 18 11 Temporal decision (recent vs.new stimuli)

5000

Garraux et al. (2005) fMRI 11 1 Paced finger tapping (timing vs. order) 2000

Jahanshahi et al. (2006)(Exp. B)

PET 8 14 Temporal reproduction (long vs.short intervals)

2000

Jech et al. (2005) fMRI 9 4 Temporal reproduction (increasein duration length)

500-16800

Kawashima et al. (2000) fMRI 8 4 Paced finger tapping (paced vs.cued finger)

1500

Knutson et al. (2004) fMRI 19 7 Temporal order (script vs.chronological order)

2500

Larsson et al. (1996) fMRI 8 4 Paced finger tapping (paced vs.cued finger)

3800

Lejeune et al. (1997) PET 12 4 Paced finger tapping (paced vs.cued finger)

2700

Lewis & Miall (2002) fMRI 8 9 Temporal reproduction (reproductionvs. control pressing)

3000

Macar et al. (2002) PET 15 6 Temporal reproduction (temporal vs.control pressing)

2200–3200–9000-13000

Macar et al. (2004) fMRI 13 2 Temporal reproduction 2500

Nani et al. 5

Page 7: AperTO - Archivio Istituzionale Open Access dell ...

Table 1. (continued )

Study Tool Participants FociExperimental Task

(Contrast or Condition) Duration (msec)

(temporal vs. force)

Murai & Yotsumoto(2016) (Exp. B)

fMRI 21 16 Temporal duration reproduction(supra vs. subsecond)

2500

Rekkas et al. (2005)(Exp. A)

fMRI 10 3 Temporal discrimination (semantictemporal order vs. spatial location)

6000

Rekkas et al. (2005)(Exp. B)

fMRI 10 5 Temporal discrimination (autobiographictemporal order vs. spatial location)

6000

Riecker et al. (2003) fMRI 8 8 Paced finger tapping (isochronous trainsof clicks at different frequencies)

6000

Riecker et al. (2006) fMRI 8 13 Paced finger tapping (self-paced syllablerepetition at different frequencies)

6000

Rubia et al. (1998) fMRI 6 4 Paced finger tapping (high- vs.low-frequency synchronization)

3000

Shergill et al. (2006) fMRI 8 9 Temporal articulation (paced vs. cued) 2000

Stevens et al. (2007) fMRI 31 12 Paced finger tapping (syncopationvs. synchronization)

750–1500–3500

Taniwaki et al. (2006) fMRI 12 16 Temporal execution (self-initiated vs.externally triggered movements)

40000

Teki & Griffiths (2016) fMRI 19 9 Temporal reproduction (temporalvs. control)

1500–2000

Tracy et al. (2000) fMRI 6 3 Temporal production (durationreproduction at different intervals)

1200–2400

Wittmann, Simmons,et al. (2010) (Exp. A)

fMRI 14 3 Temporal reproduction (temporalvs. control)

3000

Wittmann, Simmons,et al. (2010) (Exp. B)

fMRI 14 7 Temporal reproduction (temporalvs. control)

9000

Wittmann, Simmons,et al. (2010) (Exp. C)

fMRI 14 3 Temporal reproduction (temporalvs. control)

18000

Wittmann et al.(2011) (Exp. A)

fMRI 27 8 Temporal reproduction (temporalvs. control)

3000

Wittmann et al.(2011) (Exp. B)

fMRI 27 4 Temporal reproduction (temporalvs. control)

9000

Wittmann et al.(2011) (Exp. C)

fMRI 27 6 Temporal reproduction (temporalvs. control)

18000

Suprasecond motor timing condition: 28 articles, 34 experiments, 459 participants, and 261 foci of activation

Apaydın et al. (2018) fMRI 18 9 Temporal perception (timevs. reward)

2500

Carvalho et al.(2016) (Exp. A)

fMRI 16 16 Temporal expectation (nonperiodicvs. periodic)

1340–2000

Carvalho et al.(2016) (Exp. B)

fMRI 16 7 Temporal expectation (nonperiodicvs. periodic)

1340–2000

Coull et al. (2000) fMRI 6 18 Temporal estimation (long vs.short target presentation)

600–1400

Coull et al. (2004) fMRI 12 10 Temporal discrimination (increase 540–1080–1620

6 Journal of Cognitive Neuroscience Volume X, Number Y

Page 8: AperTO - Archivio Istituzionale Open Access dell ...

Table 1. (continued )

Study Tool Participants FociExperimental Task

(Contrast or Condition) Duration (msec)

in attention to time)

Coull et al. (2013)(Exp. B)

fMRI 27 15 Temporal prediction (temporal vs.neutral orienting)

600–1000–1400

Dirnberger et al. (2012) fMRI 7 7 Incorrect time estimation (aversivevs. neutral context)

1800–2200

Harrington et al. (2004) fMRI 24 26 Temporal duration discrimination(main effect of encoding)

1200–1800

Harrington et al. (2010) fMRI 20 7 Temporal duration discrimination(time vs. pitch)

10000–12500

Herrmann et al. (2014) fMRI 19 5 Temporal rate change perception(sound vs. silent trials)

6400

Jueptner et al. (1996) fMRI 6 6 Temporal discrimination (temporalvs. constant velocity)

1000

Lewis & Miall (2003b)(Exp. B)

fMRI 8 13 Temporal discrimination (temporalvs. length)

3000

Livesey et al. (2007) fMRI 10 4 Temporal discrimination (temporalvs. color)

1000–1500

Lux et al. (2003) fMRI 14 5 Temporal estimation (temporalsynchrony vs. spatial orientation)

1050

Morillon et al. (2009) fMRI 27 17 Temporal estimation (durationvs. color)

100–8400

Pfeuty et al. (2015) fMRI 29 6 Temporal estimation (durationvs. color)

1500

Pouthas et al. (2005) fMRI 6 6 Temporal estimation (long vs.short duration)

1300

Rao et al. (2001)(Exp. A)

fMRI 17 1 Temporal discrimination (temporalperception vs. control)

2500

Rao et al. (2001)(Exp. B)

fMRI 17 3 Temporal discrimination (temporalperception vs. control)

5000

Rao et al. (2001)(Exp. C)

fMRI 17 10 Temporal discrimination (temporalperception vs. control)

75000

Rao et al. (2001)(Exp. D)

fMRI 17 9 Temporal discrimination (temporalperception vs. control)

10000

Smith et al. (2003) fMRI 20 7 Temporal discrimination (temporalvs. order)

1000

Smith et al. (2011) fMRI 32 3 Temporal discrimination (temporalvs. order)

2100

Tsukamoto et al.(2006)

fMRI 20 13 Temporal estimation (true vs. falsefeedback condition)

3000

Ustun et al. (2017) fMRI 15 10 Temporal discrimination(time vs. visual)

2400–6000

Wiener et al. (2014) fMRI 25 17 Temporal discrimination(time vs. color)

2000

Suprasecond perceptual (nonmotor) timing condition: 22 articles, 26 experiments, 445 participants, and 250 activation foci

Nani et al. 7

Page 9: AperTO - Archivio Istituzionale Open Access dell ...

So, although interesting meta-analyses have beenpublished in the last 15 years, substantial disagreementabout the dynamics of functional activations of the brainstructures associated with time processing still persists.Furthermore, some methodological innovations supportthe choice to carry out a novel meta-analysis on timeperception, as the last works on this topic used an old ver-sion of GingerALE software (Ortuño et al., 2011; Wieneret al., 2010). In fact, after 2016 the BrainMap team re-ported some technical errors affecting this tool. In partic-ular, codes for the thresholding procedure with both falsediscovery rate and cluster-level family-wise error methodswere not working as expected, so that the results could bebiased by false positives (for a detailed discussion, seeEickhoff, Laird, Fox, Lancaster, & Fox, 2017). In particular,Eickhoff et al. (2016) have explicitly discouraged the use offalse discovery rate, in virtue of a high susceptibility tofalse positive results. These authors have recommendedthe implementation of cluster-level family-wise errorthresholding, which was used in the present work, alongwith the new GingerALE version in which the errors of thealgorithm have been corrected. Furthermore, previousmeta-analyses have worked on a number of studies thatwere far smaller than ours. Wiener et al. (2010) examined41 neuroimaging studies, Schwartze et al. (2012) exam-ined 42 fMRI studies, and Ortuño et al. (2011) examined

35 studies; in contrast, this meta-analysis examined 84studies, thus basically doubling the samples of the past.

METHODS

Selection of Studies

The MEDLINE database was queried to identify all pub-lished experiments on time perception or processing thatutilized functional neuroimaging tool (fMRI or PET).Using the search engine PubMed (https://www.ncbi.nlm.nih.gov/pubmed/), we adopted a MeSH hierarchyroutine. Furthermore, we used the function [ALL] ofPubMed to find MeSH terms in all the fields of the eligi-ble articles. The search algorithm was constructed asfollows:

(time perception OR [ALL] time estimation OR [ALL]time processing [ALL]) AND (functional magneticresonance OR [ALL] fMRI OR [ALL] positron emissiontomography OR [ALL] PET [ALL]).

Additionally, references from previous review articles (Teki,2016; Merchant, Harrington, & Meck, 2013) and meta-analyses (Schwartze et al., 2012; Ortuño et al., 2011; Wieneret al., 2010) about time perception were reviewed for inclu-sion. At the time of the selection phase (June 2018), over 170

Figure 1. PRISMA flow chart.Overview of the selectionstrategy.

8 Journal of Cognitive Neuroscience Volume X, Number Y

Page 10: AperTO - Archivio Istituzionale Open Access dell ...

full-text articles were assessed for eligibility. In particular,two expert researchers included in the meta-analysis thestudies on the basis of the following criteria:

a) articles were published in a peer-reviewed journal;b) articles included only healthy participants;c) articles adopted a whole-brain analysis (not just ROIs);d) articles reported coordinates of activation in Talairach/

Tourneaux or in Montreal Neurological Institute ste-reotactic brain space;

e) articles included at least one contrast of timing task(not contrast task-rest).

We therefore included in our meta-analysis 84 pub-lished articles, for a total of 109 experiments, 1018 coordi-nates of activation, and 1218 healthy participants (seeTable 1 for detailed information about the sample). Tofacilitate the subsequent analysis, we converted all nativeMontreal Neurological Institute coordinates into Talairach/Tourneaux brain space using the icbm2tal transformation(Laird et al., 2010).Following the previous studies of Lewis and Miall (2003a)

and Wiener et al. (2010), our data set was partitioned intotwo main dimensions: interval duration (i.e., subsecondand suprasecond) and type (i.e., motor and nonmotor) ofexperimental time estimation task. Accordingly, we carried

out our meta-analysis on four distinct samples: subsecondmotor (287 coordinates and 24 experiments), subsecondnonmotor (220 coordinates and 25 experiments), supra-second motor (261 coordinates and 34 experiments),and suprasecond nonmotor (250 coordinates and 26 ex-periments; see also Table 1). In this data set, we includedall the 41 studies considered by Wiener et al. (2010) and21 studies—also included by Wiener et al. (2010)—of 28of those considered by Lewis and Miall (2003a). Sevenstudies were excluded because one was an EEG study,one was a PET study on rhesus monkeys, four had restcontrast, and one had no dissociation between sub- andsuprasecond tasks. The entire process of data selectionwas performed using the Preferred Reporting Items forSystematic Reviews and Meta-analyses (PRISMA) State-ment international guidelines (Liberati et al., 2009; Moher,vLiberati, Tetzlaff, & Altman, 2009; for the overview ofthe selection strategy, see Figure 1).

Activation Likelihood Estimation and ModeledAlteration Creation

To statistically summarize the results of the retrievedexperiments for each of the four conditions previously iden-tified, we performed an activation likelihood estimation

Figure 2. Results of the ALE analyses for the four conditions (cluster level corrected, p < .05). Colors from light blue to red represent increasing ALEvalues. Slices on the axial plane are shown in radiological convention (i.e., right is left, left is right).

Nani et al. 9

Page 11: AperTO - Archivio Istituzionale Open Access dell ...

Table 2. Significant ALE Results for the Subsecond Motor Condition Meta-analysis

Cluster Volume (mm3) Label SideExtremaValue

Talairach

x y z

1 34,248 Medial frontal gyrus (BA 6) L 0.028 −6 −4 52

Medial frontal gyrus (BA 6) R 0.024 4 −4 60

Medial frontal gyrus (BA 6) R 0.022 8 4 56

Precentral gyrus (BA 6) L 0.022 −56 0 20

Precentral gyrus (BA 4) L 0.019 −36 −22 56

Superior frontal gyrus (BA 6) L 0.018 −2 6 48

Subgyral (BA 6) L 0.016 −24 −4 52

Middle frontal gyrus (BA 6) L 0.013 −26 −10 42

Cingulate gyrus (BA 32) R 0.012 10 10 36

Precentral gyrus (BA 6) L 0.012 −44 0 48

Insula (BA 13) L 0.011 −42 −10 16

Postcentral gyrus (BA 3) L 0.009 −48 −16 44

Superior temporal gyrus (BA 22) L 0.009 −52 6 4

Precentral gyrus (BA 6) L 0.009 −14 −16 66

Precentral gyrus (BA 44) L 0.008 −50 0 8

Superior frontal gyrus (BA 6) R 0.007 12 16 62

Precentral gyrus (BA 6) L 0.007 −48 −2 34

Superior frontal gyrus (BA 6) L 0.007 −8 8 70

Middle frontal gyrus (BA 6) L 0.007 −56 2 40

Inferior parietal lobule (BA 40) L 0.007 −40 −36 44

Inferior frontal gyrus (BA 45) L 0.006 −52 18 6

2 12,352 Ventral posterior medial nucleus (thalamus) L 0.028 −14 −20 6

Lentiform nucleus (putamen) L 0.022 −26 −8 6

Claustrum L 0.008 −30 10 2

Claustrum L 0.006 −32 −8 −8

3 11,752 Lentiform nucleus (putamen) R 0.018 20 4 12

Lentiform nucleus (putamen) R 0.017 20 −2 10

Substania nigra R 0.015 10 -20 −6

Lentiform nucleus (putamen) R 0.015 24 −12 6

Medial dorsal nucleus (thalamus) R 0.014 10 −18 6

Caudate body R 0.011 20 12 14

Claustrum R 0.008 26 14 16

Insula (BA 13) R 0.007 30 16 16

Ventral posterior lateral nucleus (thalamus) R 0.007 24 −24 10

10 Journal of Cognitive Neuroscience Volume X, Number Y

Page 12: AperTO - Archivio Istituzionale Open Access dell ...

Table 2. (continued )

Cluster Volume (mm3) Label SideExtremaValue

Talairach

x y z

4 9512 Culmen (cerebellum) R 0.019 14 −52 −18

Culmen (cerebellum) R 0.019 38 −50 −30

Declive (cerebellum) R 0.014 28 −60 −22

Tonsil (cerebellum) R 0.009 28 −42 −30

Tonsil (cerebellum) R 0.006 20 −58 −34

Declive (cerebellum) R 0.005 2 −60 −16

5 9336 Culmen (cerebellum) L 0.012 −22 -56 −24

Tonsil (cerebellum) L 0.012 −34 −42 −44

Declive (cerebellum) L 0.011 −30 -58 −18

Tonsil (cerebellum) L 0.010 −24 −58 −44

Uvula (cerebellum) L 0.010 −34 −62 −26

Tonsil (cerebellum) L 0.009 −28 -52 −34

Tonsil (cerebellum) L 0.007 −38 −58 −44

Declive (cerebellum) L 0.007 −12 −64 −18

Pyramis (cerebellum) L 0.007 −44 −64 −32

Contributors to Cluster 1

11 foci from Karabanov et al. (2009) 2 foci from Jantzen et al. (2007)

10 foci from Lewis et al. (2004) 2 foci from Jantzen et al. (2004)

6 foci from Schubotz & von Cramon (2001) (Exp. A) 2 foci from Jancke et al. (2000)

5 foci from Pope et al. (2005) 2 foci from Rao et al. (1997) (Exp. A)

4 foci from Ortuño et al. (2002) 3 foci from Rao et al. (1997) (Exp. B)

3 foci from Murai & Yotsumoto (2016) (Exp. A) 2 foci from Bueti et al. (2008) (Exp. A)

3 foci from Jantzen et al. (2005) 1 focus from Berns et al. (1997)

3 foci from Pastor et al. (2004) 1 focus from Bengtsson et al. (2005)

3 foci from Oullier et al. (2005) (Exp. A) 1 focus from Penhune et al. (1998) (Exp. B)

4 foci from Oullier et al. (2005) (Exp. B)

Contributors to Cluster 2

4 foci from Lewis et al. (2004) 1 focus from Jantzen et al. (2005)

3 foci from Mayville et al. (2002) 1 focus from Jantzen et al. (2004)

3 foci from Karabanov et al. (2009) 1 focus from Penhune et al. (1998) (Exp. A)

2 foci from Schubotz & von Cramon (2001) (Exp. A) 1 focus from Oullier et al. (2005) (Exp. A)

2 foci from Rao et al. (1997) (Exp. A) 1 focus from Oullier et al. (2005) (Exp. B)

3 foci from Rao et al. (1997) (Exp. B) 1 focus from Bueti et al. (2008) (Exp. A)

2 foci from Pope et al. (2005)

Nani et al. 11

Page 13: AperTO - Archivio Istituzionale Open Access dell ...

(ALE) as implemented in GingerAle 2.3.6 (Eickhoff, Bzdok,Laird, Kurth, & Fox, 2012; Turkeltaub et al., 2012; Eickhoffet al., 2009). The four resulting maps have been clustered ata level of p < .05, family-wise error-corrected for multiplecomparisons, with a cluster-forming threshold of p < .001(Eickhoff et al., 2016, 2017).

The ALE is a technique capable of providing informa-tion about the reliability of results by means of a compar-ison between the experiments and a sample of referencestudies from the existing literature (Laird et al., 2005). AnALE meta-analysis considers every experiment focus asbeing a Gaussian probability distribution:

p dð Þ ¼ 1

σ3ffiffiffiffiffiffiffiffiffiffiffi2πð Þ3

q e−d2

2σ2

in which d is the Euclidean distance between the focusand the voxels whereas is the standard deviation ofthe distribution.A modeled alteration map is then determined for every

experiment as the combination of the Gaussian probabil-ity distribution of every experiment focus. Subsequently,the ALE map is obtained from the union of all the mod-eled alteration maps.The ALE map is thresholded on the basis of a null

hypothesis, obtained recursively by producing all thepossible random distribution of foci. Then, to obtain acluster-level correction, a similar procedure is repeatedto define the size of nonrandom clusters of activation.The significance of values within the ALE map is finallycalculated by testing the null hypothesis that active voxelsare randomly distributed across the brain. The iterative

Table 2. (continued )

Cluster Volume (mm3) Label SideExtremaValue

Talairach

x y z

Contributors to Cluster 3

6 foci from Mayville et al. (2002) 2 foci from Oullier et al. (2005) (Exp. A)

6 foci from Karabanov et al. (2009) 2 foci from Oullier et al. (2005) (Exp. A)

2 foci from Lewis et al. (2004) 2 foci from Bueti et al. (2008) (Exp. A)

2 foci from Schubotz & von Cramon (2001) (Exp. A) 2 foci from Jantzen et al. (2004)

2 foci from Pope et al. (2005)

Contributors to Cluster 4

3 foci from Bengtsson et al. (2005) 1 focus from Rao et al. (1997) (Exp. A)

3 foci from Karabanov et al. (2009) 1 focus from Rao et al. (1997) (Exp. B)

2 foci from Lewis et al. (2004) 1 focus from Pope et al. (2005)

2 foci from Mayville et al. (2002) 1 focus from Oullier et al. (2005) (Exp. A)

1 focus from Jantzen et al. (2004) 1 focus from Ortuño et al. (2002)

1 focus from Schubotz & von Cramon (2001) (Exp. A) 1 focus from Bueti et al. (2008) (Exp. A)

Contributors to Cluster 5

2 foci from Schubotz & von Cramon (2001) (Exp. A) 1 focus from Pope et al. (2005)

2 foci from Karabanov et al. (2009) 1 focus from Penhune et al. (1998) (Exp. A)

2 foci from Aso et al. (2010) (Exp. A) 1 focus from Penhune et al. (1998) (Exp. B)

1 focus from Lewis et al. (2004) 1 focus from Oullier et al. (2005) (Exp. A)

1 focus from Jantzen et al. (2007) 1 focus from Oullier et al. (2005) (Exp. B)

1 focus from Jantzen et al. (2005) 1 focus from Ortuño et al. (2002)

1 focus from Jantzen et al. (2004) 1 focus from Bueti et al. (2008) (Exp. A)

For each cluster brain regions, cluster size (mm3), side (hemisphere), ALE value, Talairach coordinates, and contributors to cluster (foci) wereprovided. BA = Brodmann’s area; L = left; R = right.

12 Journal of Cognitive Neuroscience Volume X, Number Y

Page 14: AperTO - Archivio Istituzionale Open Access dell ...

Table 3. Significant ALE Results for the Subsecond Perceptual Condition Meta-analysis

Cluster Volume (mm3) Label Side Extrema Value

Talairach

x y z

1 25,192 Claustrum L 0.031 −30 −6 12

Inferior frontal gyrus (BA 47) L 0.022 −48 12 0

Claustrum L 0.020 −28 24 4

Precentral gyrus (BA 44) L 0.019 −46 8 12

Lentiform nucleus (putamen) L 0.017 −22 2 8

Precentral gyrus (BA 6) L 0.016 −46 0 32

Middle frontal gyrus (BA 6) L 0.015 −24 −8 50

Inferior frontal gyrus (BA 9) L 0.015 −52 10 26

Insula (BA 13) L 0.014 −40 14 6

Lentiform nucleus (putamen) L 0.014 −24 10 4

Precentral gyrus (BA 6) L 0.009 −48 −2 8

Middle frontal gyrus (BA 6) L 0.009 −38 2 46

Middle frontal gyrus (BA 9) L 0.008 −48 22 32

Middle frontal gyrus (BA 46) L 0.008 −46 22 20

Middle frontal gyrus (BA 6) L 0.007 −48 4 48

Middle frontal gyrus (BA 9) L 0.006 −46 32 38

2 14,896 Insula (BA 13) R 0.032 34 18 6

Caudate body R 0.022 14 8 6

Insula (BA 13) R 0.014 44 10 16

Superior temporal gyrus (BA 22) R 0.011 50 12 0

Lentiform nucleus (globus pallidus) R 0.010 20 −2 −6

3 13,208 Cingulate gyrus (BA 32) L 0.040 0 28 34

Superior frontal gyrus (BA 6) R 0.023 2 10 50

Medial frontal gyrus (BA 6) L 0.016 −2 −4 54

Cingulate gyrus (BA 32) L 0.008 −12 6 40

Medial frontal gyrus (BA 6) L 0.008 −4 −18 52

Contributors to Cluster 1

6 foci from Tomasi et al. (2015) 2 foci from Shih et al. (2009)

5 foci from Schubotz & von Cramon (2001) (Exp. B) 2 foci from Bueti et al. (2008)

4 foci from Lewis & Miall (2003b) 1 focus from Wittmann et al. (2010a) (Exp. A)

4 foci from Ferrandez et al. (2003) 1 focus from Wittmann et al. (2010a) (Exp. C)

3 foci from Tregellas et al. (2006) 1 focus from van Wassenhove et al. (2011) (Exp. A)

3 foci from Schubotz et al. (2000) 1 focus from van Wassenhove et al. (2011) (Exp. B)

2 foci from Beudel et al. (2009) (Exp. A) 1 focus from Coull & Nobre (1998) (Exp. A)

Nani et al. 13

Page 15: AperTO - Archivio Istituzionale Open Access dell ...

process generates a vector that can be considered as ahistogram representing the null distribution. Voxel-levelthresholds can be obtained from this. The cluster-levelthreshold is obtained through the calculation of the clus-ter size distribution, after 1000 permutations (Acar,Seurinck, Eickhoff, & Moerkerke, 2018).

Correlation between Maps and Voxels’Contribution to Correlation Analysis

To compare the four different conditions, we computed allthe possible correlations between the maps. To do so, wecomputed the linear correlation, by means of Pearson’s r,between the two matrices containing the voxel values ofthe map associated with each condition.

For every condition, we further analyzed the voxel-wise contribution to the correlation. This is an innovativemapping analysis that allows us to appreciate the mar-ginal contribution of each voxel in the correlation of twomaps, so as to better identify the voxels exerting moreinfluence in the correlation. Specifically, to examine theregional differences between the four conditions, we ap-plied the analysis leave-one-voxel-out so as to evaluatethe voxels’ contribution to the correlation (VCC) betweenthe maps (Mancuso et al., 2019). Every pair of correspond-ing voxels vi1 and vi2 was removed from both maps, andafter every removal, the Pearson’s r between the two mapswas recalculated. The difference rGLOBAL – RViREMOVED be-tween the original value of r and the value obtained afterremoving the couple of voxels vi1 and vi2 was attributed tovoxel vi. This procedure was repeated for vj, vk, …; then,

Table 3. (continued )

Cluster Volume (mm3) Label Side Extrema Value

Talairach

x y z

2 foci from Bengtsson et al. (2009) (Exp. A) 1 focus from Beudel et al. (2009) (Exp. B)

2 foci from Tipples et al. (2013) 1 focus from Aso et al. (2010) (Exp. B)

Contributors to Cluster 2

5 foci from Tomasi et al. (2015) 1 focus from Shih et al. (2010)

3 foci from Lewis & Miall (2003b) 1 focus from Ferrandez et al. (2003)

3 foci from Tipples et al. (2013) 1 focus from Beudel et al. (2009) (Exp. A)

3 foci from Schubotz & von Cramon (2001) (Exp. B) 2 focus from Beudel et al. (2009) (Exp. B)

2 foci from Tregellas et al. (2006) 1 focus from Maquet et al. (1996)

2 foci from Schubotz et al. (2000) 1 focus from Bueti et al. (2008) (Exp. B)

1 focus from Wittmann et al. (2010a) (Exp. B) 1 focus from Aso et al. (2010) (Exp. B)

1 focus from van Wassenhove et al. (2011) (Exp. B)

Contributors to Cluster 3

3 foci from Bengtsson et al. (2009) (Exp. A) 1 focus from van Wassenhove et al. (2011) (Exp. B)

2 foci from Lewis & Miall, 2003b 1 focus from Tomasi et al. (2015)

2 foci from Ferrandez et al. (2003) 1 focus from Gutyrchik et al. (2010)

2 foci from Tipples et al. (2013) 1 focus from Beudel et al. (2009) (Exp. A)

2 foci from Shih et al. (2009) 1 focus from Beudel et al. (2009) (Exp. B)

2 foci from Schubotz & von Cramon (2001) (Exp. B) 1 focus from Tregellas et al. (2006)

1 focus from Wittmann et al. (2010a) (Exp. A) 1 focus from Schubotz et al. (2000)

1 focus from Wittmann et al. (2010a) (Exp. B) 1 focus from Maquet et al. (1996)

1 focus from van Wassenhove et al. (2011) (Exp. A)

For each cluster brain regions, cluster size (mm3), side (hemisphere), ALE value, Talairach coordinates, and contributors to cluster (foci) wereprovided. BA = Brodmann’s area; L = left; R = right.

14 Journal of Cognitive Neuroscience Volume X, Number Y

Page 16: AperTO - Archivio Istituzionale Open Access dell ...

Table 4. Significant ALE Results for the Suprasecond Motor Condition Meta-analysis

Cluster Volume(mm3) Label SideExtremaValue

Talairach

x Y Z

1 19,008 Superior frontal gyrus (BA 6) L 0.024 −2 10 50

Medial frontal gyrus (BA 6) R 0.020 2 0 62

Cingulate gyrus (BA 32) R 0.018 6 20 36

Medial frontal gyrus (BA 6) L 0.011 −10 −22 46

Cingulate gyrus (BA 32) L 0.009 −6 24 34

Medial frontal gyrus (BA 8) L 0.009 −4 28 42

Paracentral lobule (BA 31) L 0.007 −2 −24 42

2 11,784 Inferior frontal gyrus (BA 44) R 0.023 52 8 22

Insula (BA 13) R 0.020 42 12 6

Inferior frontal gyrus (BA 47) R 0.014 36 20 -4

Claustrum R 0.010 30 14 6

Inferior frontal gyrus (BA 45) R 0.009 46 28 4

Superior temporal gyrus (BA 22) R 0.005 56 8 4

3 6824 Mammillary body L 0.020 −12 −18 4

Lentiform nucleus (putamen) L 0.013 −18 −6 12

Caudate body L 0.010 −14 8 12

Lateral posterior nucleus (thalamus) L 0.008 −14 −18 18

Contributors to Cluster 1

3 foci from Dudukovic & Wagner (2007) 1 focus from Wittmann et al. (2010b) (Exp. B)

3 foci from Stevens et al. (2007) 1 focus from Wittmann et al. (2010b) (Exp. C)

2 foci from Lewis & Miall (2002) 1 focus from Murai & Yotsumoto (2016)

2 foci from Larsson et al. (1996) 1 focus from Lejeune et al. (1997)

2 foci from Taniwaki et al. (2006) 1 focus from Jech et al. (2005)

2 foci from Rubia et al. (1998) 1 focus from Coull et al. (2013) (Exp. A)

2 foci from Riecker et al. (2003) 1 focus from Basso et al. (2003)

2 foci from Macar et al. (2002) 1 focus from Shergill et al. (2006)

1 focus from Wittmann et al. (2011) (Exp. A) 1 focus from Riecker et al. (2006)

1 focus from Wittmann et al. (2011) (Exp. B) 1 focus from Rekkas et al. (2005) (Exp. B)

1 focus from Wittmann et al. (2011) (Exp. C) 1 focus from Macar et al. (2004)

Contributors to Cluster 2

2 foci from Wittmann et al. (2010) (Exp. B) 1 focus from Wittmann et al. (2010b) (Exp. C)

2 foci from Coull et al. (2013) (Exp. A) 1 focus from Murai & Yotsumoto (2016)

2 foci from Ackermann et al. (2001) 1 focus from Lewis & Miall (2002)

2 foci from Shergill et al. (2006) 1 focus from Lejeune et al. (1997)

Nani et al. 15

Page 17: AperTO - Archivio Istituzionale Open Access dell ...

the difference values were normalized, so as to obtain avoxel-wise estimation of the local contribution to the globalcorrelation.

Negative VCC values denote the voxels whose removalproduces an increase of the correlation, which meansthat they do not positively contribute to the global r. Incontrast, positive VCC values denote the voxels contrib-uting to a positive correlation between the two maps,whose removal determines an rGLOBAL decrease. TheVCC can therefore provide us with the information aboutthe regions that contribute to the similarity and the dis-similarity of the two maps.

Hierarchical Clustering and Features ofRelevance Estimation

To better investigate the differences and commonalitiesbetween the various conditions, we applied a hierarchicalclustering analysis using as entry elements the four ob-tained ALE maps. As a first step, the peak ALE valuesfor each of the ROIs were calculated. Parcels were iden-tified on the basis of the Talairach Atlas. The 113 ROIsthat showed nonzero peak values at least in one of thefour conditions were passed to subsequent analyses,performed using the Orange Canvas software. An N ×M matrix (4 conditions × 113 ROIs) was used as input.Euclidean distances were computed between rows (i.e.,between conditions), and hierarchical clustering wasperformed with the Ward’s linkage method. We thenused the information gain (IG) criterion (Alhaj, Siraj,Zainal, Elshoush, & Elhaj, 2016) to identify the mostinformative brain regions capable of distinguishingbetween the four conditions. The IG is based on a

measure of entropy and is commonly used in the fea-ture-ranking procedures, so as to identify those featuresthat are particularly relevant for the accurate implemen-tation of a given task (e.g., to assign an item to the cor-rect class among many).

Reliability Analysis

To investigate the reliability of the obtained results, weapplied the “fail-safe” technique. This kind of approach,frequently used in classical meta-analyses, was first in-troduced by Rosenthal (1979), and recently, Acar et al.(2018) made it suitable for ALEmeta-analyses. This methodaddresses the issue of the so-called “file-drawer effect,” thatis, the smaller chance for a study to show nonstatisti-cally significant results to get published. This bias couldlead to overestimated effects in meta-analyses because ofthe lack of contraevidences in the pool of consideredstudies. The procedure that we used, proposed by Acar,Seurinck, Eickhoff, and Moerkerke (2017; https://github.com/NeuroStat/GenerateNull), allows to introduce “noiseexperiments” to the meta-analysis pool. The injection ofgrowing levels of noise raises the statistical threshold, al-lowing to test how many unpublished experiments shouldexist to nullify the obtained results. More in details, theAcar’s algorithm observes the distributions of sample sizeand amount of reported foci per experiment in the pool ofreal experiments. These two parameters are pivotal for theimplementation of the ALE method and the identificationof the effect (Eickhoff et al., 2016). Noise experiments arethen created randomly by combining values of sample sizeand amount of reported foci per experiment taken fromthe respective real distributions (Acar et al., 2017). The

Table 4. (continued )

Cluster Volume(mm3) Label SideExtremaValue

Talairach

x Y Z

1 focus from Wittmann et al. (2011) (Exp. A) 1 focus from Brunia et al. (2000)

1 focus from Wittmann et al. (2011) (Exp. B) 1 focus from Basso et al. (2003)

1 focus from Wittmann et al. (2011) (Exp. C) 1 focus from Taniwaki et al. (2006)

1 focus from Wittmann et al. (2010b) (Exp. A) 1 focus from Stevens et al. (2007)

Contributors to Cluster 3

3 foci from Riecker et al. (2006) 1 focus from Coull et al. (2013) (Exp. A)

2 foci from Taniwaki et al. (2006) 1 focus from Ackermann et al. (2001)

2 foci from Riecker et al. (2003) 1 focus from Stevens et al. (2007)

1 focus from Teki et al. (2016) 1 focus from Shergill et al. (2006)

For each cluster brain regions, cluster size (mm3), side (hemisphere), ALE value, Talairach coordinates, and contributors to cluster (foci) wereprovided. BA = Brodmann’s area; L = left; R = right.

16 Journal of Cognitive Neuroscience Volume X, Number Y

Page 18: AperTO - Archivio Istituzionale Open Access dell ...

Table 5. Significant ALE Results for the Suprasecond Perceptual Condition Meta-analysis

Cluster Volume (mm3) Label SideExtremaValue

Talairach

x y z

1 19,240 Insula (BA 13) R 0.037 34 16 6

Inferior frontal gyrus (BA 44) R 0.025 50 10 20

Middle frontal gyrus (BA 9) R 0.016 36 32 32

Middle frontal gyrus (BA 10) R 0.011 32 44 2

Middle frontal gyrus (BA 46) R 0.010 44 40 26

Middle frontal gyrus (BA 6) R 0.009 30 16 50

Inferior frontal gyrus (BA 46) R 0.008 44 32 6

Middle frontal gyrus (BA 9) R 0.008 44 22 28

Precentral gyrus (BA 6) R 0.008 38 2 32

Inferior frontal gyrus (BA 45) R 0.007 60 24 12

2 16,848 Medial frontal gyrus (BA 32) L 0.030 −4 8 46

Medial frontal gyrus (BA 6) L 0.030 −6 4 52

Superior frontal gyrus (BA 6) R 0.024 6 10 54

Cingulate gyrus (BA 32) R 0.013 12 28 26

Medial frontal gyrus (BA 8) R 0.012 2 22 46

Middle frontal gyrus (BA 6) R 0.012 30 4 62

Medial frontal gyrus (BA 9) R 0.011 2 34 30

Superior frontal gyrus (BA 9) R 0.011 22 40 30

Superior frontal gyrus (BA 6) R 0.010 22 6 60

Superior frontal gyrus (BA 6) R 0.009 16 16 54

Middle frontal gyrus (BA 6) R 0.008 30 2 52

3 16,016 Claustrum L 0.033 −34 14 4

Inferior frontal gyrus (BA 44) L 0.028 −48 6 20

Lentiform nucleus (putamen) L 0.014 −14 8 −4

Lentiform nucleus (putamen) L 0.011 −16 4 4

Middle frontal gyrus (BA 9) L 0.009 −54 8 36

Middle frontal gyrus (BA 9) L 0.008 −40 14 28

Precentral gyrus (BA 44) L 0.006 −50 12 4

4 7368 Inferior parietal lobule (BA 40) R 0.020 48 −42 42

Inferior parietal lobule (BA 40) R 0.019 40 −48 42

Insula (BA 13) R 0.009 52 −34 20

Superior temporal gyrus (BA 13) R 0.009 52 −42 22

Inferior parietal lobule (BA 40) R 0.009 54 −32 46

Superior temporal gyrus (BA 41) R 0.009 54 −28 10

Superior temporal gyrus (BA 22) R 0.005 54 −38 10

Nani et al. 17

Page 19: AperTO - Archivio Istituzionale Open Access dell ...

Table 5. (continued )

Cluster Volume (mm3) Label SideExtremaValue

Talairach

x y z

Contributors to Cluster 1

6 foci from Lewis & Miall (2003b) 2 foci from Rao et al. (2001) (Exp. C)

4 foci from Carvalho et al. (2016) (Exp. A) 1 focus from Wiener et al. (2014)

4 foci from Smith et al. (2003) 1 focus from Livesey et al. (2007)

3 foci from Pfeuty et al. (2015) 1 focus from Harrington et al. (2004)

3 foci from Morillon et al. (2009) 1 focus from Coull et al. (2004)

3 foci from Coull et al. (2013) (Exp. B) 1 focus from Apaydın et al. (2018)

2 foci from Ustun et al. (2017) 1 focus from Rao et al. (2001) (Exp. A)

2 foci from Jueptner et al. (1996) 1 focus from Rao et al. (2001) (Exp. B)

2 foci from Coull et al. (2000) 1 focus from Pouthas et al. (2005)

Contributors to Cluster 2

3 foci from Coull et al. (2004) 1 focus from Pfeuty et al. (2015)

3 foci from Carvalho et al. (2016)(Exp. A) 1 focus from Lewis & Miall (2003b)

2 foci from Wiener et al. (2014) 1 focus from Herrmann et al. (2014)

2 foci from Morillon et al. (2009) 1 focus from Harrington et al. (2010)

2 foci from Harrington et al. (2004) 1 focus from Coull et al. (2013)(Exp. B)

2 foci from Coull et al. (2000) 1 focus from Apaydın et al. (2018)

2 foci from Smith et al. (2003) 1 focus from Rao et al. (2001) (Exp. A)

2 foci from Pouthas et al. (2005) 1 focus from Rao et al. (2001) (Exp. B)

1 focus from Ustun et al. (2017) 1 focus from Rao et al. (2001) (Exp. C)

1 focus from Smith et al. (2011) 1 focus from Rao et al. (2001) (Exp. D)

Contributors to Cluster 3

3 foci from Lewis & Miall (2003b) 2 foci from Tsukamoto et al. (2006)

3 foci from Harrington et al. (2010) 2 foci from Rao et al. (2001) (Exp. C)

3 foci from Carvalho et al. (2016) (Exp. A) 1 focus from Smith et al. (2011)

3 foci from Rao et al. (2001) (Exp. D) 1 focus from Morillon et al. (2009)

2 foci from Ustun et al. (2017) 1 focus from Herrmann et al. (2014)

2 foci from Livesey et al. (2007) 1 focus from Harrington et al. (2004)

2 foci from Coull et al. (2013) (Exp. B) 1 focus from Coull et al. (2004)

2 foci from Coull et al. (2000) 1 focus from Rao et al. (2001) (Exp. B)

Contributors to Cluster 4

2 foci from Harrington et al. (2004) 1 focus from Lewis & Miall (2003b)

2 foci from Carvalho et al. (2016) (Exp. B) 1 focus from Coull et al. (2013) (Exp. B)

2 foci from Tsukamoto et al. (2006) 1 focus from Coull et al. (2004)

18 Journal of Cognitive Neuroscience Volume X, Number Y

Page 20: AperTO - Archivio Istituzionale Open Access dell ...

final step is the repetition of the ALE analysis using asinput the real pool injected with variable amount ofnoise experiments.In this study, the range of injected noise was between k

4and 3k, where k is the number of real experiments. In eachrepetition, we added the k

4 of noise (i.e., 12 ALE maps).Finally, we created a map showing the residual real contri-bution of each brain area after an increasing injection ofnoise. Because the sample size and the amount of reportedfoci per experiment are specific for each pool used for ameta-analysis, we repeated the fail-safe procedure for eachof the four conditions investigated in this study.

RESULTS

The Patterns of Brain Areas Involved inTime Perception

The ALE analysis of the subsecond motor conditionshows activations mainly localized in medial and superiorfrontal gyri, left precentral gyrus, right cingulate gyrus,left superior temporal gyrus, left inferior parietal lobule,insula, right claustrum, thalamus, putamen, right caudatebody, substantia nigra, and cerebellum (Figure 2 andTable 2).The ALE analysis of the subsecond perceptual con-

dition shows activations mainly localized in the left claus-trum, left middle, medial and inferior frontal gyri, leftprecentral gyrus, left putamen, insula, right caudate body,right superior temporal gyrus, right globus pallidus, leftcingulate gyrus, and right superior frontal gyrus (Figure 2and Table 3).

The ALE analysis of the suprasecond motor conditionshows activations mainly localized in left superior frontalgyrus, medial frontal gyrus, cingulate gyrus, right inferiorfrontal gyrus, right insula, right claustrum, right superiortemporal gyrus, left mammillary body, left putamen, leftcaudate body, and left thalamus (Figure 2 and Table 4).

The ALE analysis of the suprasecond perceptual condi-tion shows activations mainly localized in right insula,right middle and inferior frontal gyri, precentral gyrus,left medial frontal gyrus, right superior frontal gyrus, rightcingulate gyrus, right medial frontal gyrus, left claustrum,left inferior frontal gyrus, left putamen, left middle frontalgyrus, right inferior parietal lobule, and right superiortemporal gyrus (Figure 2 and Table 5).

Correlation between Maps and Voxels’Contribution to Correlation Analysis

With regard to the correlation analyses, the highest value(r = .51) was found between subperceptual and supra-perceptual conditions (Table 6). The correlation betweensupramotor versus supraperceptual was r = .38, followedby submotor versus supramotor (r = .36) and submotorversus subperceptual (r = .23). In light of this, the sub-motor condition is more correlated with the supramotorthan the subperceptual. The subperceptual condition ismore correlated with the supraperceptual than the sub-motor. The supramotor condition is more correlatedwith the supraperceptual than the submotor. Finally,the supraperceptual condition is more correlated withthe subperceptual than the supramotor.

Table 5. (continued )

Cluster Volume (mm3) Label SideExtremaValue

Talairach

x y z

1 focus from Ustun et al. (2017) 1 focus from Coull et al. (2000)

1 focus from Smith et al. (2011) 1 focus from Apaydın et al. (2018)

1 focus from Morillon et al. (2009)

For each cluster brain regions, cluster size (mm3), side (hemisphere), ALE value, Talairach coordinates, and contributors to cluster (foci) wereprovided. BA = Brodmann’s area; L = left; R = right.

Table 6. Correlation Values (Pearson’s r) between Each Couple of Conditions

Subsecond motor Subsecond perceptual Suprasecond motor Suprasecond perceptual

Subsecond motor 0.23 0.36 0.18

Subsecond perceptual 0.37 0.51

Suprasecond motor 0.38

Suprasecond perceptual

Nani et al. 19

Page 21: AperTO - Archivio Istituzionale Open Access dell ...

The VCC analysis produced the following results. Thevoxels driving the correlation between the subsecondmotor and subsecond perceptual conditions are mainlylocalized in the left superior and middle frontal gyrus, leftsuperior insula, right caudate body, left precentral gyrus,and right lateral globus pallidus (Figure 3).

The voxels driving the correlation between the sub-second motor and suprasecond motor conditions aremainly localized in the bilateral medial frontal gyrus,the left ventral posterior lateral nucleus of the thalamus,and the left putamen (Figure 3).

The voxels driving the correlation between the sub-second perceptual and the suprasecond perceptual con-ditions are mainly localized in the right superior andmedial frontal gyrus, bilateral anterior insula, left inferiorfrontal gyrus, and left putamen (Figure 3).

The voxels driving the correlation between the suprase-cond motor and the suprasecond perceptual conditions

are mainly localized in the right superior, medial and infe-rior frontal gyri, and left putamen (Figure 3).

Hierarchical Clustering and Features ofRelevance Estimation

The dendrogram obtained through the hierarchical clus-tering shows the highest similarity between the two per-ceptual conditions (i.e., sub- and suprasecond). Thesuprasecond motor condition results to be more similarto the perceptual cluster than to the subsecond motorcondition (Figure 4). This configuration is coherent withthe correlation results, because the highest value wasfound between the two perceptual conditions. The su-prasecond motor condition shows a slightly higher corre-lation with the suprasecond perceptual condition ratherthan with the subsecond motor one. The IG identifiesthe left BA 31, left and right putamen, right BA 6, right

Figure 3. Results of the contribution leave-one-voxel-out to the correlation analyses. Colors from light blue to red represent increasing contributionto the overall correlation between the two conditions (values were normalized between 0 and 1). Values of r are the results of the correlation beforethe leave-one-out procedure. Slices on the axial plane are shown in radiological convention (i.e., right is left, left is right).

20 Journal of Cognitive Neuroscience Volume X, Number Y

Page 22: AperTO - Archivio Istituzionale Open Access dell ...

BA 32, left BA 45, left BA 8, right ventral anterior nucleusof the thalamus, left lateral and medial globus pallidus,right BA 44, left posterior and anterior BA 13, left BA22, left BA 47, left caudate body, left BA 43, left BA 4,and right BA 24 as the most informative areas for discrim-inating between conditions. Of note, none of these brainregions was active in only one of the four conditions.The overlap of the patterns of activations obtained

with the ALE maps associated with the four conditionsidentifies the SMA as the area in which conditions showthe greatest overlap (100%; Figure 5).As the SMA appears to be the area with the greatest

overlap between conditions, we have examined how itsactivation can vary in relation to sub- and suprasecondtasks, as well as to motor and nonmotor tasks (Figure 6).

Reliability

The fail-safe analysis shows a good resistance of the re-sults (Figure 7). Most of the activation blobs, across thefour conditions, survive after the 50% level of noise is in-jected. Of note, the SMA is still present even after the

Figure 4. Results of thehierarchical clustering.Dendrogram was built using theWard’s linkage. The nodes’values represent dissimilarity.

Figure 5. Overlap between the ALE maps associated to each of the fourconditions. Colors from light blue to red represent increasing overlapvalues. Voxels showing value equal to 100% were found to be activethrough all the four conditions. Slices on the axial plane are shown inradiological convention (i.e., right is left, left is right).

Nani et al. 21

Page 23: AperTO - Archivio Istituzionale Open Access dell ...

injection of the 300% of noise in three of the four condi-tions (the exception is the subsecond perceptual).Subcortical regions show a good resistance, especiallywith regard to the subsecond conditions (motor andnonmotor), where they survive above the 50% of thenoise level. A similar trend is observed for the insula aswell.

Overall, the fail-safe analysis about the robustness ofour results is reassuring, as most of them have main-tained significance after noise injections of 50% and theSMA even after an injection of 300% of noise.

DISCUSSION

Our meta-analytical investigation, which used a new cor-rected version for the ALE and an innovative fail-safetechnique, mostly confirms the findings of the previousmeta-analyses. However, patterns of activation relatedto subsecond and suprasecond tasks are much more sim-ilar than previously thought. Both subsecond and supra-second tasks recruit cortical and subcortical areas, but

subcortical areas contribute more to the pattern asso-ciated with subsecond tasks than to the pattern associ-ated with suprasecond tasks, which instead receivesmore contributions from cortical activations. This is inline with the idea that distinct timing mechanisms mayoperate at different timescales (Buhusi & Meck, 2005), aswell as to the idea that intervals below the 1-sec range aresupposed to be more dependent on sensory and automaticprocesses, whereas the detection and estimation of intervalslonger than 1 sec are thought to relymore on cognitive func-tions (Lewis & Miall, 2003a; Rammsayer, 1999). Several cog-nitive processes are in fact likely to contribute to theperception/construction of time (Meck, 2005; Pouthas &Perbal, 2004). For instance, the degree of attention canmod-ulate and influence the estimate of time (Coull, Vidal, Nazarian,&Macar, 2004; Nobre&O’Reilly, 2004). Also, workingmemoryand executive processes come into play in time judgments(Radua, Del Pozo, Gomez, Guillen-Grima, & Ortuño, 2014).In tasks when the participant is required to compare thedifferent durations of two stimuli, the duration of the firststimulus must be held in the STM storage so as to decidewhich stimulus is longer (Livesey, Wall, & Smith, 2007).

Figure 6. Sagittal and axialsections of the ALE in all thefour conditions in which thedifferent activation profile of theSMA can be appreciated. Colorsfrom light blue to red representincreasing ALE values. Slices onthe axial plane are shown inradiological convention (i.e.,right is left, left is right).

22 Journal of Cognitive Neuroscience Volume X, Number Y

Page 24: AperTO - Archivio Istituzionale Open Access dell ...

Motor and perceptual conditions have similar profiles,but the pattern associated with the motor condition ismore extended in subcortical and parietal sites than thepattern associated with the perceptual condition. Therecruitment of parietal sites is probably due to the well-known functions of these cortical areas in the elaborationof the where and how of spatial vision (Oliveri et al.,2009; Alexander, Cowey, & Walsh, 2005), which are fun-damental for building spatiotemporal representationsand then voluntarily orienting the motor response (butsee further on for a more detailed discussion of thispoint).Overall, our analysis provides evidence that the pat-

terns associated with the four conditions are character-ized by many common activations in cerebral sites; inparticular, they share activations in the following regions:the insula, superior, medial and inferior frontal gyri, pre-central gyrus, cingulate gyrus, superior temporal gyrus,claustrum, putamen, and caudate body. This various setof brain areas highlights the importance of the interplaybetween cortical and subcortical components, especiallyof frontal insular areas and the striatum, in performingtasks related to time (Wittmann, 2013; Wittmann & vanWassenhove, 2009).The insular cortex is thought to integrate different

signals originating from the outside world and within the or-ganism, a function that is fundamental for building an

interoceptive feeling of the state of the body (Critchley,Wiens, Rotshtein, Ohman, & Dolan, 2004; Craig, 2002).This integration would rely on an accumulation mechanismof physiological changes in the posterior insula (Wittmann,Simmons, Aron, & Paulus, 2010) and then proceed in sequen-tial representations of homeostatic states in a posterior-to-anterior progression (Craig, 2009b; Singer, Critchley,& Preuschoff, 2009). This would result in a unified metar-epresentation capable of forming our internal feelingsmoment after moment, the development of which wouldbe at the basis of our inner experience of duration (Craig,2009a).

Undoubtedly, the great involvement of the frontal cor-tex (superior, medial, middle, and inferior gyri) in all thefour conditions points, as already suggested, to the rele-vance of higher order cognitive processes, such as atten-tion and working memory, to perform time estimationtasks, which in terms of cognitive resources are quitedemanding (Radua et al., 2014). In particular, the SMA(which corresponds to parts of the superior and medialfrontal gyri) has been variously associated with both sen-sory and sensorimotor processing of temporal intervals,with the rostral SMA (or pre-SMA) more engaged insensory, nonsequential and suprasecond temporal process-ing, and the caudal SMA (or SMA-proper) more engaged insensorimotor, sequential and subsecond temporal process-ing (Schwartze et al., 2012). Recently a study by Protopapa

Figure 7. Results of the fail-safe analyses for the four conditions. Colors from light blue to red represent increasing resistance to injected noise. Sliceson the axial plane are shown in radiological convention (i.e., right is left, left is right).

Nani et al. 23

Page 25: AperTO - Archivio Istituzionale Open Access dell ...

et al. (2019) has suggested that the entire SMA may codifychronomaps in a progressive way, with the spatial progres-sion following a typical rostro-caudal direction, with anterioractivations (those of pre-SMA) more associated with theprocessing of shorter time periods, and posterior activa-tions (those of SMA-proper) more associated with the pro-cessing of longer time periods. As can be seen by the SMAactivations illustrated by Figure 7, our results further sup-port the rostrocaudal gradient proposed by Schwartzeet al. (2012), who associate the activity of the pre-SMAand SMA-proper not only to the elaboration of different du-ration of time intervals but also to different tasks. Our re-sults provide evidence for this view, as the gradient in theactivation of the SMA appears to be associated not only withthe length of the interval but also with the nature of thetasks, as caudal activations seem to be more related tothe motor condition whereas rostral activations seem tobe more related to the nonmotor (or perceptual) condi-tion. This finding is also in line with what was reportedby Wiener et al. (2011).

The activation of the precentral gyrus, especially the leftone, may be related to information rehearsal (Rao et al.,2001) and directed attention (Tzourio et al., 1997). As al-ready observed by Wiener et al. (2010), this area would beinvolved in maintaining task-directed attention. Further-more, parts of the precentral gyrus may differ with respectto functional specificity: The equivalent location of the handarea on the motor homunculus may be associated with sub-second tasks, whereas the equivalent location of the mouthsite may be associated with suprasecond tasks. In particular,subsecond tasks might engage the precentral gyrus in real(or imagined) timed hand movements (Oullier, Jantzen,Steinberg, & Kelso, 2005), whereas suprasecond tasks mightengage this area in vocal (or subvocal) counting strategies(Hinton, Harrington, Binder, Durgerian, & Rao, 2004).

The cingulate cortex, in turn, lies at the crossroads ofemotion, sensation, and action. In particular, the ACC islargely connected with the anterior insula, and both areasare essential parts of the salience network (Seeley et al.,2007); therefore, it might support the implementation ofattentional and executive control in temporal tasks.

As already seen, the striatum has been repeatedly re-lated to time perception. In virtue of its many connec-tions with cortical regions, this subcortical structurewould play the role of an “internal clock” capable of inte-grating the oscillatory cortical activity. More specifically,the striatal beat frequency (SBF) model proposed byMatell and Meck (2000, 2004) suggests that synchronizedfrequencies of large cortical areas are the neural inputsconstituting the time code for constructing temporalrepresentations. According to the SBF model, the dopa-minergic system gives a burst of dopamine to the striatum,thus providing the onset signal for timing a certain stimulus(Meck et al., 2008). The spiny neurons of the striatumwould then detect the concurrent activity of specific beatpatterns of cortical fluctuations. Our results provide furthersupport for the SBF model. Contributions of the striatum

are in fact present in all the activation patterns associatedwith the four conditions of temporal processing.Activity in the superior temporal gyrus has been re-

lated to temporal discrimination and reproduction of au-ditory stimuli ( Jahanshahi, Jones, Dirnberger, & Frith,2006; Rao et al., 2001; Sakai et al., 1999), as well as tothe generation of rhythms learned through the auditorymodality (Bengtsson, Ehrsson, Forssberg, & Ullen, 2005;Lewis, Wing, Pope, Praamstra, & Miall, 2004; Jancke,Loose, Lutz, Specht, & Shah, 2000; Rao et al., 1997).This region, therefore, plays an important role in thetiming of auditory stimuli (Bueti, van Dongen, & Walsh,2008). However, it has also reported activity in this areain the temporal discrimination of visual stimuli (Coullet al., 2004; Ferrandez et al., 2003). As suggested by theauthors of these studies and given that the evaluation oftemporal intervals is more difficult for visual than forauditory stimuli (Westheimer, 1999), activation of the su-perior temporal gyrus may support an auditory imagerystrategy or subvocalization, which could help discrimi-nate visual durations.An interesting finding that has not been reported by

previous meta-analyses about time perception, with theexception of Wiener et al. (2012), is the contribution ofthe claustrum. Activations of this region, whose exactfunction still remains an unresolved mystery, are presentin every pattern associated with the four experimentalconditions. The claustrum has many distributed connec-tions to practically all the cortical areas (frontal, premo-tor, motor, ventral temporal, entorhinal, ventral anteriorcingulate, visual, somatosensory, and olfactory), as well asto important subcortical structures (lateral amygdala, pu-tamen, caudate nucleus, and globus pallidus; Baizer,Sherwood, Noonan, & Hof, 2014; Mathur, 2014;Torgerson & Van Horn, 2014). In virtue of these exten-sive projections, Crick and Koch (2005) proposed thatthe claustrum might integrate different sensations intoa unitary conscious representation. This proposal hasbeen further developed by Smythies, Edelstein, andRamachandran (2012, 2014), who put forward a detailedhypothesis of how the claustrum might bring about thistemporal integration. According to this hypothesis, theclaustrum may serve as a “synchrony detector,” thus ca-pable of elaborating higher order temporal patterns tocoordinate and bind information coming from all overacross the brain. Of note, our analysis shows that theclaustrum has significant activations in the perceptualconditions. This provides further evidence that the claus-trum might play an important role in constructing thetemporal infrastructure of consciousness or, as we preferto call it, the property of consciousness of being a “tem-poral glue” capable of binding different percepts in a co-herent conscious scene. If the claustrum is to be seen ashaving this role, there seems to be in the brain tworegions that are supposed to be “synchrony detectors”for temporal patterns (i.e., the claustrum and the SBF).We do not think that this situation is contradictory, but

24 Journal of Cognitive Neuroscience Volume X, Number Y

Page 26: AperTO - Archivio Istituzionale Open Access dell ...

rather, we think that data can be reconciled if we con-sider, as we have noted, that the claustrum is much moreactivated in perceptual (and cognitive) tasks whereas thestriatum appears to be more involved in motor tasks. Thetwo “synchrony detectors” may, therefore, work in syn-ergy with the prevalence of either one or the other ac-cording to the nature and demands of the current task.Another difference with previous meta-analyses is the

absence of the cerebellum in three of the activation pat-terns associated with the temporal conditions. In paststudies, the cerebellum has been related to the elabora-tion of temporal information (Mangels, Ivry, & Shimizu,1998; Jueptner et al., 1995; Ivry, 1993; Ivry & Keele,1989). More recently, the cerebellum has been linked tothe processing of duration-based rather than beat-based in-tervals (Cope, Grube, Singh, Burn, & Griffiths, 2014; Grube,Cooper, Chinnery, & Griffiths, 2010). Furthermore, it hasbeen proposed that it could play a role in the neural net-work capable of timing short intervals, as it has been foundto be more involved in subsecond rather than in suprase-cond tasks (Fierro et al., 2007; Koch et al., 2007; Lee et al.,2007). The label-based meta-analysis carried out by Lewisand Miall (2003a) has confirmed the involvement of cere-bellum in time processing. In particular, the authors foundthe activation of the right cerebellum in subsecond tasksand the activation of the left cerebellum in suprasecondtasks. The review of Penney and Vaitilingam (2008), whichalso maintained the division of experiments between sub-second and suprasecond categories, found that the cere-bellum was the most frequently activated structure insubsecond timing tasks. But as pointed out by Wieneret al. (2010), this review included activations that weresubtracted both from control tasks and from the rest con-dition, thus making it difficult to distinguish between ac-tivations related to timing procedures or to othercognitive processes. Finally, Wiener et al. (2010) reportedthe cerebellum to be active in the subsecond categoryonly. Our results confirm in part the findings of thesemeta-analyses. As already said, activations of the cerebel-lum are totally absent in three of the four conditionstaken into examination here, save for the subsecond mo-tor. A possible explanation for this result is that the othermeta-analyses have probably used less restrictive statisti-cal thresholds. In fact the cerebellum was clearly presentonly in our unthresholded ALE maps, which suggests thatbeyond a certain statistical threshold its activation be-comes nonsignificant. On the other hand, this resultmay be more likely due to the large sample of studiesused in this meta-analysis. Although all the selected stud-ies conducted whole-brain analyses, not all of them re-ported activity in the cerebellum. More specifically, theactivation foci of the cerebellum found by experimentswere as follows: For subsecond motor, 19 (79%) reportedactivation foci, and 5 (21%) did not. For subsecond per-ceptual, 13 (52%) reported activation foci, and 12 (48%)did not. For suprasecond motor, 13 (38%) reported acti-vation foci, and 21 (62%) did not. Finally, for suprasecond

perceptual, 9 (35%) reported activation foci, and 17(65%) did not. In light of these data, the cerebellar ac-tivity does not seem to be so relevant as to contributeeffectively to the processing of timing tasks that do notpertain to the subsecond motor condition.

Another result of this meta-analysis that contrasts withprevious findings is the activation of the parietal cortex.Lewis and Miall (2003a) found this region activated in su-prasecond tasks, whereas Wiener et al. (2010) reportedactivations of this area (especially the inferior parietallobe) in subsecond motor, subsecond perceptual andsuprasecond motor tasks, but not in suprasecond percep-tual tasks. Contrary to Lewis and Miall (2003a) and inpartial agreement with Wiener et al. (2010), we found ac-tivations of the parietal cortex in subsecond and motortasks only. The activations were mainly restricted to theinferior parietal lobule, with the exception of the post-central gyrus. These findings support the idea that theparietal cortex, in particular the inferior parietal lobule,could play a role in estimating time; however, what itdoes exactly remains unclear. Leon and Shadlen (2003)found activation of the posterior parietal cortex (i.e., lat-eral inferior lobule) of macaques in subsecond tasks. Onthe other hand, Oliveri et al. (2009) applied TMS on theparietal cortex to test healthy participants and patientswith right brain damage during a task requiring settingthe midpoint of a time suprasecond interval. The authorsfound that inhibition of the right posterior parietal cortexinduces a directional bias in the time bisection task. Inturn, Vicario, Martino, and Koch (2013) applied to twogroups of healthy adults a transcranial direct currentstimulation (anodal, cathodal, or sham) over the leftand right posterior parietal cortices and compared theperformance of the groups in suprasecond time repro-duction tasks. Results showed that, when cathodal stim-ulation was applied to the right posterior parietal cortex,the temporal accuracy was affected and led participantsto overestimate time intervals, whereas when it was ap-plied to the left posterior parietal cortex, the variabilityin reproducing temporal intervals was reduced. Theauthors reported no effect for the anodal stimulation.As functions of the parietal cortex are intimately asso-ciated with attentional direction and spatial processing,it is extremely difficult to separate these components intiming tasks. The hope is that in the future better devisedexperimental protocols could help distinguish the contri-butions of parietal regions to time processing on the onehand and to spatial and attentional processes on theother.

The correlation analysis between the four experimentalcategories reveals that the strongest couple is subsecondperceptual and suprasecond perceptual, followed by su-prasecond motor and suprasecond perceptual, sub-second motor and suprasecond motor, and subsecondmotor and subsecond perceptual. In other words, thepatterns of activations associated with the perceptualconditions exhibit a greater similarity than the patterns

Nani et al. 25

Page 27: AperTO - Archivio Istituzionale Open Access dell ...

of activations associated with the motor conditions. Theseresults further suggest that the frontal components, whichare predominant in the perceptual conditions, are morespecific for the nature of the task rather than for time du-ration. Furthermore, the subdivision between motor andperceptual processing may be due to the fact that thesetwo conditions rely on different functional patterns offrontal sites, which in part have common activations, asit is also suggested by the VCC analysis, which shows thata variety of frontal voxels contribute to the correlations.

The hierarchical clustering analysis confirms the cor-relation results, as it shows that the highest similarity isbetween sub- and suprasecond perceptual conditions.Also the result that the suprasecond motor condition isslightly more similar with the suprasecond perceptualcondition is in line with the correlation analysis (accord-ing to which the more correlated second couple is supra-second motor and suprasecond perceptual). This mayindicate that the similarity is mostly guided by the frontalcomponents, which are predominant in the suprasecondperceptual condition.

The IG analysis gives us the areas in which we can ob-serve the major differences in peak activation betweenthe four conditions. Interestingly, 13 of 19 areas are local-ized in the left hemisphere, and among those, most arefrontal regions, which suggests that part of the left frontalpattern may have the role of discriminating tasks in eachcondition.

Finally, different from the meta-analysis of Wiener et al.(2010), which found only two areas (i.e., the SMA and theinferior frontal gyrus) with significant activations in a con-junction analysis of all the timing conditions, our overlapof the patterns of activations obtained with the ALE mapsassociated with the four conditions reveals just the SMAas the area of 100% of overlap. This provides further ev-idence for the pivotal role of the SMA in estimating timeintervals, independent of the condition.

The SMA, along subcortical structures and the insula, arethe areas that survive better the fail-safe procedure. Thesebrain regions, therefore, might be considered as core net-work nodes for time perception, thus suggesting that oursense of time could emerge from the complex interactionof higher order and automatic cognitive processes.

Limitations and Future Directions

A concern with the ALE approach, which is, however, oneof the most applied methods in coordinate-based meta-analyses, is that results might be driven by few experi-ments, thus representing a particular case among theones selected for the meta-analysis rather that an overalldescriptive view. Still, it should be observed that a mini-mum amount of 20 experiments is generally consideredto be sufficient to avoid this issue (Eickhoff et al., 2016),so that analyses on large databases, as the one carried outhere, should not be biased. Also, our sample is very wellbalanced, as we performed our analysis on a comparable

sample size for each condition. However, although wewere able to identify a considerable number of studies,others could have been omitted.Future studies are needed to unravel the brain mech-

anisms at the root of our perception/construction oftime. We need especially better protocols so as to distin-guish in more details the roles played by cerebral struc-tures in temporal processing. For instance, the activity ofprimary sensorimotor areas might be merely due to realor imaginary counting strategies. It is therefore funda-mental to understand in what degree the areas respon-sible for sensory perception and motor ideation andexecution contribute to the estimation of time per se.Finally, whether or not the cerebellum should be consid-ered an essential brain structure in time processing is stilla moot point, as it does not show concordance acrossstudies. Our results suggest that its involvement is asso-ciated with the subsecond motor but not with the otherconditions. Further analyses are needed to address thisinteresting issue.

Conclusion

The variability of time perception models may be ac-counted for by the peculiar nature of temporal phe-nomena, which are not directly related to objects in theoutside world and for which we do not possess a dedi-cated sensory organ. This meta-analysis provides evidencefor the existence of specific patterns of activations that canbe associated with the perception and/or estimation oftime in subsecond and suprasecond motor and nonmotortasks. These patterns, however, are not radically differentbut have in part similar profiles of activations, as they sharethe contributions of common brain areas. Motor and sub-second tasks recruit slightly more contributions of subcor-tical structures than perceptual and suprasecond tasks,which in turn show patterns of activations more associatedwith cortical areas. All the conditions, however, exhibit astrong component of frontal areas and, in particular, a ro-bust activation of the SMA. Rostral and caudal parts of thisarea play a pivotal role in discriminating not only differenttime intervals but also in relation to the nature of the taskconditions. The SMA, in addition with the striatum (espe-cially the putamen) and the claustrum, might constitute acore circuit in the different networks engaged when thesense of time is constructed by the brain.

Acknowledgments

This study was supported by the Fondazione Carlo Molo andFondazione Sanpaolo (F. Cauda, PI), Turin. A. N. conceivedthe experiment, implemented data collection, drafted, and re-vised the article. J. M. analyzed the data, drafted, and revisedthe article. L. D. implemented data collection, drafted, and re-vised the article. S. D. revised the article. T. C. devised the toolsof analysis, supervised the analysis of data, drafted, and revisedthe article. F. C. supervised the analysis of data, drafted, andrevised the article.

26 Journal of Cognitive Neuroscience Volume X, Number Y

Page 28: AperTO - Archivio Istituzionale Open Access dell ...

Reprint requests should be sent to Tommaso Costa, Departmentof Psychology, Via Po 14, 10123 Turin, Italy, or via e-mail: [email protected].

REFERENCES

Acar, F., Seurinck, R., Eickhoff, S. B., & Moerkerke, B. (2018).Assessing robustness against potential publication bias inactivation likelihood estimation (ALE) meta-analyses for fMRI.PLOS ONE, 13, e0208177.

Ackermann, H., Riecker, A., Mathiak, K., Erb, M., Grodd, W.,& Wildgruber, D. (2001). Rate-dependent activation of aprefrontal-insular-cerebellar network during passive listeningto trains of click stimuli: An fMRI study. NeuroReport, 12,4087–4092.

Alexander, I., Cowey, A., & Walsh, V. (2005). The right parietalcortex and time perception: Back to Critchley and theZeitraffer phenomenon. Cognitive Neuropsychology, 22,306–315.

Alhaj, T. A., Siraj, M. M., Zainal, A., Elshoush, H. T., & Elhaj, F.(2016). Feature selection using information gain forimproved structural-based alert correlation. PLOS ONE, 11,e0166017.

Apaydın, N., Üstün, S., Kale, E. H., Çelikağ, İ., Özgüven, H. D.,Baskak, B., et al. (2018). Neural mechanisms underlying timeperception and reward anticipation. Frontiers in HumanNeuroscience, 12, 115.

Aso, K., Hanakawa, T., Aso, T., & Fukuyama, H. (2010). Cerebro-cerebellar interactions underlying temporal informationprocessing. Journal of Cognitive Neuroscience, 22,2913–2925.

Baizer, J. S., Sherwood, C. C., Noonan, M., & Hof, P. R. (2014).Comparative organization of the claustrum: What doesstructure tell us about function? Frontiers in SystemsNeuroscience, 8, 117.

Basso, G., Nichelli, P., Wharton, C. M., Peterson, M., & Grafman,J. (2003). Distributed neural systems for temporalproduction: A functional MRI study. Brain ResearchBulleting, 59, 405–411.

Battelli, L., Pascual-Leone, A., & Cavanagh, P. (2007). The‘when’ pathway of the right parietal lobe. Trends in CognitiveSciences, 11, 204–210.

Bengtsson, S. L., Ehrsson, H. H., Forssberg, H., & Ullen, F.(2005). Effector-independent voluntary timing: Behaviouraland neuroimaging evidence. European Journal ofNeuroscience, 22, 3255–3265.

Bengtsson, S. L., Ullén, F., Ehrsson, H. H., Hashimoto, T., Kito,T., Naito, E., et al. (2009). Listening to rhythms activatesmotor and premotor cortices. Cortex, 45, 62–71.

Berns, G. S., Cohen, J. D., & Mintun, M. A. (1997). Brain regionsresponsive to novelty in the absence of awareness. Science,276, 1272–1275.

Beudel, M., Renken, R., Leenders, K. L., & de Jong, B. M.(2009). Cerebral representations of space and time.Neuroimage, 44, 1032–1040.

Botzung, A., Denkova, E., & Manning, L. (2008). Experiencingpast and future personal events: Functional neuroimagingevidence on the neural bases of mental time travel. Brainand Cognition, 66, 202–212.

Brunia, C. H. M., de Jong, B. M., van den Berg-Lenssen, M. M.C., & Paans, A. M. J. (2000). Visual feedback about timeestimation is related to a right hemisphere activationmeasured by PET. Experimental Brain Research, 130,328–337.

Bueti, D., van Dongen, E. V., & Walsh, V. (2008). The role ofsuperior temporal cortex in auditory timing. PLOS ONE, 3,e2481.

Bueti, D., Walsh, V., Frith, C., & Rees, G. (2008). Different braincircuits underlie motor and perceptual representations oftemporal intervals. Journal of Cogntive Neuroscience, 20,204–214.

Buhusi, C. V., & Meck, W. H. (2005). What makes us tick?Functional and neural mechanisms of interval timing. NatureReviews Neuroscience, 6, 755–765.

Buonomano, D. V., & Maass, W. (2009). State-dependentcomputations: Spatiotemporal processing in corticalnetworks. Nature Reviews Neuroscience, 10, 113–125.

Carvalho, F. M., Chaim, K. T., Sanchez, T. A., & de Araujo, D. B.(2016). Time-perception network and default modenetwork are associated with temporal prediction in aperiodic motion task. Frontiers in Human Neuroscience,10, 268.

Cope, T. E., Grube, M., Singh, B., Burn, D. J., & Griffiths, T. D.(2014). The basal ganglia in perceptual timing: Timingperformance in multiple system atrophy and huntington’sdisease. Neuropsychologia, 52, 73–81.

Coull, J. T., Davranche, K., Nazarian, B., & Vidal, F. (2013).Functional anatomy of timing differs for production versusprediction of time intervals. Neuropsychologia, 51,309–319.

Coull, J. T., Frith, C. D., Büchel, C., & Nobre, A. C. (2000).Orienting attention in time: Behavioural andneuroanatomical distinction between exogenous andendogenous shifts. Neuropsychologia, 38, 808–819.

Coull, J. T., & Nobre, A. C. (1998). Where and when to payattention: The neural systems for directing attention tospatial locations and to time intervals as revealed byboth PET and fMRI. Journal of Neuroscience, 18,7426–7435.

Coull, J. T., & Nobre, A. (2008). Dissociating explicit timing fromtemporal expectation with fMRI. Current OpinionNeurobiology, 18, 137–144.

Coull, J. T., Vidal, F., Nazarian, B., & Macar, F. (2004). Functionalanatomy of the attentional modulation of time estimation.Science, 303, 1506–1508.

Craig, A. D. (2002). How do you feel? Interoception: The senseof the physiological condition of the body. Nature ReviewsNeuroscience, 3, 655–666.

Craig, A. D. (2009a). Emotional moments across time: Apossible neural basis for time perception in the anteriorinsula. Philosophical Transactions of the Royal Society ofLondon: Series B: Biologucal Sciences, 364, 1933–1942.

Craig, A. D. (2009b). How do you feel—Now? The anteriorinsula and human awareness. Nature Reviews Neuroscience,10, 59–70.

Crick, F. C., & Koch, C. (2005). What is the function of theclaustrum? Philosophical Transactions of the Royal Societyof London: Series B: Biologucal Sciences, 360, 1271–1279.

Critchley, H. D., Wiens, S., Rotshtein, P., Ohman, A., & Dolan,R. J. (2004). Neural systems supporting interoceptiveawareness. Nature Neuroscience, 7, 189–195.

Dirnberger, G., Hesselmann, G., Roiser, J. P., Preminger, S.,Jahanshahi, M., & Paz, R. (2012). Give it time: Neuralevidence for distorted time perception and enhancedmemory encoding in emotional situations. Neuroimage,63, 591–599.

Dudukovic, N. M., & Wagner, A. D. (2007). Goal-dependentmodulation of declarative memory: Neural correlates oftemporal recency decisions and novelty detection.Neuropsychologia, 45, 2608–2620.

Eickhoff, S. B., Bzdok, D., Laird, A. R., Kurth, F., & Fox, P. T.(2012). Activation likelihood estimation meta-analysisrevisited. Neuroimage, 59, 2349–2361.

Eickhoff, S. B., Laird, A. R., Fox, P. M., Lancaster, J. L., &Fox, P. T. (2017). Implementation errors in the GingerALE

Nani et al. 27

Page 29: AperTO - Archivio Istituzionale Open Access dell ...

software: Description and recommendations. HumanBrain Mapping, 38, 7–11.

Eickhoff, S. B., Laird, A. R., Grefkes, C., Wang, L. E., Zilles, K., &Fox, P. T. (2009). Coordinate-based activation likelihoodestimation meta-analysis of neuroimaging data: A random-effects approach based on empirical estimates of spatialuncertainty. Human Brain Mapping, 30, 2907–2926.

Eickhoff, S. B., Nichols, T. E., Laird, A. R., Hoffstaedter, F.,Amunts, K., Fox, P. T., et al. (2016). Behavior, sensitivity, andpower of activation likelihood estimation characterized bymassive empirical simulation. Neuroimage, 137, 70–85.

Ferrandez, A. M., Hugueville, L., Lehéricy, S., Poline, J. B.,Marsault, C., & Pouthas, V. (2003). Basal ganglia andsupplementary motor area subtend duration perception: AnfMRI study. Neuroimage, 19, 1532–1544.

Fierro, B., Palermo, A., Puma, A., Francolini, M., Panetta, M. L.,Daniele, O., et al. (2007). Role of the cerebellum in timeperception: A TMS study in normal subjects. Journal of theNeurological Science, 263, 107–112.

Garraux, G., McKinney, C., Wu, T., Kansaku, K., Nolte, G., &Hallett, M. (2005). Shared brain areas but not functionalconnections controlling movement timing and order.Journal of Neuroscience, 25, 5290–5297.

Gibbon, J., Malapani, C., Dale, C. L., & Gallistel, C. (1997).Toward a neurobiology of temporal cognition: Advances andchallenges. Current Opinion in Neurobiology, 7, 170–184.

Grube, M., Cooper, F. E., Chinnery, P. F., & Griffiths, T. D.(2010). Dissociation of duration-based and beat-basedauditory timing in cerebellar degeneration. Proceedings ofthe National Academy of Sciences, U.S.A., 107, 11597–11601.

Gutyrchik, E., Churan, J., Meindl, T., Bokde, A. L., vonBernewitz, H., Born, C., et al. (2010). Functionalneuroimaging of duration discrimination on two differenttime scales. Neuroscience Letters, 469, 411–415.

Harrington, D. L., Boyd, L. A., Mayer, A. R., Sheltraw, D. M., Lee,R. R., Huang, M., et al. (2004). Neural representation ofinterval encoding and decision making. Brain Research:Cognitive Brain Research, 21, 193–205.

Harrington, D. L., & Haaland, K. Y. (1999). Neuralunderpinnings of temporal processing: A review of focallesion, pharmacological, and functional imaging research.Reviews in Neuroscience, 10, 91–116.

Harrington, D. L., Zimbelman, J. L., Hinton, S. C., & Rao,S. M. (2010). Neural modulation of temporal encoding,maintenance, and decision processes. Cerebral Cortex,20, 1274–1285.

Herrmann, B., Henry, M. J., Scharinger, M., & Obleser, J. (2014).Supplementary motor area activations predict individualdifferences in temporal-change sensitivity and its illusorydistortions. Neuroimage, 101, 370–379.

Hinton, S. C., Harrington, D. L., Binder, J. R., Durgerian, S., &Rao, S. M. (2004). Neural systems supporting timing andchronometric counting: An fMRI study. Brain Research:Cognitive Brain Research, 21, 183–192.

Ivry, R. B. (1993). Cerebellar involvement in the explicitrepresentation of temporal information. Annals of theNew York Academy of Sciences, 682, 214–230.

Ivry, R. B., & Keele, S. W. (1989). Timing functions of thecerebellum. Journal Cognitive Neuroscience, 1, 136–152.

Ivry, R. B., & Schlerf, J. E. (2008). Dedicated and intrinsicmodels of time perception. Trends in Cognitive Sciences,12, 273–280.

Ivry, R. B., Spencer, R. M., Zelaznik, H. N., & Diedrichsen, J.(2002). The cerebellum and event timing. Annals of NewYork Academy of Science, 978, 302–317.

Jahanshahi, M., Jones, C. R., Dirnberger, G., & Frith, C. D.(2006). The substantia nigra pars compacta and temporalprocessing. Journal of Neuroscience, 26, 12266–12273.

Jancke, L., Loose, R., Lutz, K., Specht, K., & Shah, N. J. (2000).Cortical activations during paced finger-tapping applyingvisual and auditory pacing stimuli. Brain Research: CognitiveBrain Research, 10, 51–66.

Jantzen, K. J., Oullier, O., Marshall, M., Steinberg, F. L., &Kelso, J. A. (2007). A parametric fMRI investigation ofcontext effects in sensorimotor timing and coordination.Neuropsychologia, 45, 673–684.

Jantzen, K. J., Steinberg, F. L., & Kelso, J. A. (2004). Brainnetworks underlying human timing behavior are influencedby prior context. Proceedings of the National Academy ofSciences, U.S.A., 101, 6815–6820.

Jantzen, K. J., Steinberg, F. L., & Kelso, J. A. (2005). FunctionalMRI reveals the existence of modality and coordination-dependent timing networks. Neuroimage, 25, 1031–1042.

Jech, R., Dusek, P., Wackermann, J., & Vymazal, J. (2005).Cumulative blood oxygenation-level-dependent signalchanges support the ‘time accumulator’ hypothesis.NeuroReport, 16, 1467–1471.

Jueptner, M., Flerich, L., Weiller, C., Mueller, S. P., & Diener,H. C. (1996). The human cerebellum and temporalinformation processing—Results from a PET experiment.NeuroReport, 7, 2761–2765.

Jueptner, M., Rijntjes, M., Weiller, C., Faiss, J. H., Timmann,D., Mueller, S. P., et al. (1995). Localization of a cerebellartiming process using PET. Neurology, 45, 1540–1545.

Karabanov, A., Blom, O., Forsman, L., & Ullén, F. (2009). Thedorsal auditory pathway is involved in performance of bothvisual and auditory rhythms. Neuroimage, 44, 480–488.

Kawashima, R., Okuda, J., Umetsu, A., Sugiura, M., Inoue, K.,Suzuki, K., et al. (2000). Human cerebellum plays animportant role in memory-timed finger movement: An fMRIstudy. Journal of Neurophysiology, 83, 1079–1087.

Knutson, K. M., Wood, J. N., & Grafman, J. (2004). Brainactivation in processing temporal sequence: An fMRI study.Neuroimage, 23, 1299–1307.

Koch, G., Oliveri, M., Torriero, S., Salerno, S., Lo Gerfo, E., &Caltagirone, C. (2007). Repetitive TMS of cerebelluminterferes with millisecond time processing. ExperimentalBrain Research, 179, 291–299.

Laird, A. R., Fox, P. M., Price, C. J., Glahn, D. C., Uecker, A. M.,Lancaster, J. L., et al. (2005). ALE meta-analysis: Controllingthe false discovery rate and performing statistical contrasts.Human Brain Mapping, 25, 155–164.

Laird, A. R., Robinson, J. L., McMillan, K. M., Tordesillas-Gutiérrez, D., Moran, S. T., Gonzales, S. M., et al. (2010).Comparison of the disparity between Talairach and MNIcoordinates in functional neuroimaging data: Validation ofthe Lancaster transform. Neuroimage, 51, 677–683.

Larsson, J., Gulyas, B., & Roland, P. E. (1996). Cortical representationof self-paced finger movement. NeuroReport, 7, 463–468.

Lee, K. H., Egleston, P. N., Brown, W. H., Gregory, A. N.,Barker, A. T., & Woodruff, P. W. (2007). The role of thecerebellum in subsecond time perception: Evidence fromrepetitive transcranial magnetic stimulation. JournalCognitive Neuroscience, 19, 147–157.

Lejeune, H., Maquet, P., Bonnet, M., Casini, L., Ferrara, A.,Macar, F., et al. (1997). The basic pattern of activation inmotor and sensory temporal tasks: Positron emissiontomography data. Neuroscience Letters, 235, 21–24.

Leon, M. I., & Shadlen, M. N. (2003). Representation of time byneurons in the posterior parietal cortex of the macaque.Neuron, 38, 317–327.

Lewis, P. A., & Miall, R. C. (2002). Brain activity during non-automatic motor production of discrete multi-secondintervals. NeuroReport, 13, 1731–1735.

Lewis, P. A., & Miall, R. C. (2003a). Distinct systems forautomatic and cognitively controlled time measurement:

28 Journal of Cognitive Neuroscience Volume X, Number Y

Page 30: AperTO - Archivio Istituzionale Open Access dell ...

Evidence from neuroimaging. Current OpinionNeurobiology, 13, 250–255.

Lewis, P. A., & Miall, R. C. (2003b). Brain activation patternsduring measurement of sub- and suprasecond intervals.Neuropsychologia, 41, 1583–1592.

Lewis, P. A., & Miall, R. C. (2006a). Remembering the time:A continuous clock. Trends in Cognitive Sciences, 10,401–406.

Lewis, P. A., & Miall, R. C. (2006b). A right hemisphericprefrontal system for cognitive time measurement.Behavioural Processes, 71, 226–234.

Lewis, P. A., Wing, A. M., Pope, P. A., Praamstra, P., & Miall, R. C.(2004). Brain activity correlates differentially with increasingtemporal complexity of rhythms during initialisation,synchronisation, and continuation phases of paced fingertapping. Neuropsychologia, 42, 1301–1312.

Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gøtzsche,P. C., Ioannidis, J. P., et al. (2009). The PRISMA statement forreporting systematic reviews and meta-analyses of studiesthat evaluate health care interventions: Explanation andelaboration. Journal Clinical Epidemiology, 62, e1–e34.

Livesey, A. C., Wall, M. B., & Smith, A. T. (2007). Timeperception: Manipulation of task difficulty dissociates clockfunctions from other cognitive demands. Neuropsychologia,45, 321–331.

Lutz, K., Specht, K., Shah, N. J., & Jancke, L. (2000). Tappingmovements according to regular and irregular visualtiming signals investigated with fMRI. NeuroReport, 11,1301–1306.

Lux, S., Marshall, J. C., Ritzl, A., Zilles, K., & Fink, G. R. (2003).Neural mechanisms associated with attention to temporalsynchrony versus spatial orientation: An fMRI study.Neuroimage, 20(Suppl. 1), S58–S65.

Macar, F., Anton, J. L., Bonnet, M., & Vidal, F. (2004). Timingfunctions of the supplementary motor area: An event-relatedfMRI study. Brain Research: Cognitive Brain Research, 21,206–215.

Macar, F., Coull, J., & Vidal, F. (2006). The supplementarymotor area in motor and perceptual time processing:fMRI studies. Cognitive Processing, 7, 89–94.

Macar, F., Lejeune, H., Bonnet, M., Ferrara, A., Pouthas, V.,Vidal, F., et al. (2002). Activation of the supplementarymotor area and of attentional networks during temporalprocessing. Experimental Brain Research, 142, 475–485.

Mancuso, L., Costa, T., Nani, A., Manuello, J., Liloia, D., Gelmini,G., et al. (2019). The homotopic connectivity of thefunctional brain: A meta-analytic approach. Scientic Reports,9, 3346.

Mangels, J. A., Ivry, R. B., & Shimizu, N. (1998). Dissociablecontributions of the prefrontal and neocerebellar cortex totime perception. Brain Research: Cognitive Brain Research,7, 15–39.

Maquet, P., Lejeune, H., Pouthas, V., Bonnet, M., Casini, L.,Macar, F., et al. (1996). Brain activation induced by estimationof duration: A PET study. Neuroimage, 3, 119–126.

Matell, M. S., & Meck, W. H. (2000). Neuropsychologicalmechanisms of interval timing behavior. BioEssays,22, 94–103.

Matell, M. S., & Meck, W. H. (2004). Cortico-striatal circuits andinterval timing: Coincidence detection of oscillatoryprocesses. Brain Research: Cognitive Brain Research,21, 139–170.

Mathiak, K., Hertrich, I., Grodd, W., & Ackermann, H. (2004).Discrimination of temporal information at the cerebellum:Functional magnetic resonance imaging of nonverbalauditory memory. Neuroimage, 21, 154–162.

Mathur, B. N. (2014). The claustrum in review. Frontiers inSystems Neuroscience, 8, 48.

Mauk, M. D., & Buonomano, D. V. (2004). The neural basis oftemporal processing. Annual Review of Neuroscience, 27,307–340.

Mayville, J. M., Jantzen, K. J., Fuchs, A., Steinberg, F. L., & Kelso,J. A. (2002). Cortical and subcortical networks underlyingsyncopated and synchronized coordination revealed usingfMRI. Functional magnetic resonance imaging. Human BrainMapping, 17, 214–229.

Meck, W. H. (2005). Neuropsychology of timing and timeperception. Brain and Cognition, 58, 1–8.

Meck, W. H., Penney, T. B., & Pouthas, V. (2008). Cortico-striatalrepresentation of time in animals and humans. CurrentOpinion in Neurobiology, 18, 145–152.

Merchant, H., Harrington, D. L., & Meck, W. H. (2013). Neuralbasis of the perception and estimation of time. AnnualReview of Neuroscience, 36, 313–336.

Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009).Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Journal of ClinicalEpidemiology, 62, 1006–1012.

Morillon, B., Kell, C. A., & Giraud, A. L. (2009). Three stages andfour neural systems in time estimation. Journal ofNeuroscience, 29, 14803–14811.

Murai, Y., & Yotsumoto, Y. (2016). Context-dependent neuralmodulations in the perception of duration. Frontiers inIntegrative Neuroscience, 10, 12.

Neufang, S., Fink, G. R., Herpertz-Dahlmann, B., Willmes, K., &Konrad, K. (2008). Developmental changes in neuralactivation and psychophysiological interaction patterns ofbrain regions associated with interference control and timeperception. Neuroimage, 43, 399–409.

Nobre, A. C., & O’Reilly, J. (2004). Time is of the essence.Trends in Cognitive Sciences, 8, 387–389.

Oliveri, M., Koch, G., Salerno, S., Torriero, S., Lo Gerfo, E., &Caltagirone, C. (2009). Representation of time intervals inthe right posterior parietal cortex: Implications for a mentaltime line. Neuroimage, 46, 1173–1179.

Ortuño, F., Guillen-Grima, F., Lopez-Garcia, P., Gomez, J., &Pla, J. (2011). Functional neural networks of time perception:Challenge and opportunity for schizophrenia research.Schizophrenia Research, 125, 129–135.

Ortuño, F., Ojeda, N., Arbizu, J., Lopez, P., Martí-Climent, J. M.,Peñuelas, I., et al. (2002). Sustained attention in a countingtask: Normal performance and functional neuroanatomy.Neuroimage, 17, 411–420.

Oullier, O., Jantzen, K. J., Steinberg, F. L., & Kelso, J. A. (2005).Neural substrates of real and imagined sensorimotorcoordination. Cerebral Cortex, 15, 975–985.

Pastor, M. A., Day, B. L., Macaluso, E., Friston, K. J., &Frackowiak, R. S. (2004). The functional neuroanatomy oftemporal discrimination. Journal of Neuroscience, 24,2585–2591.

Penhune, V. B., Zattore, R. J., & Evans, A. C. (1998).Cerebellar contributions to motor timing: A PET study ofauditory and visual rhythm reproduction. Journal ofCognitive Neuroscience, 10, 752–765.

Pfeuty, M., Dilharreguy, B., Gerlier, L., & Allard, M. (2015).fMRI identifies the right inferior frontal cortex as the brainregion where time interval processing is altered by negativeemotional arousal. Human Brain Mapping, 36, 981–995.

Pope, P., Wing, A. M., Praamstra, P., & Miall, R. C. (2005). Forcerelated activations in rhythmic sequence production.Neuroimage, 27, 909–918.

Pouthas, V., George, N., Poline, J. B., Pfeuty, M.,Vandemoorteele, P. F., Hugueville, L., et al. (2005). Neuralnetwork involved in time perception: An fMRI studycomparing long and short interval estimation. Human BrainMapping, 25, 433–441.

Nani et al. 29

Page 31: AperTO - Archivio Istituzionale Open Access dell ...

Pouthas, V., & Perbal, S. (2004). Time perception dependson accurate clock mechanisms as well as unimpairedattention and memory processes. Acta NeurobiologiaeExperimentalis, 64, 367–385.

Protopapa, F., Hayashi, M. J., Kulashekhar, S., van der Zwaag,W., Battistella, G., Murray, M. M., et al. (2019). Chronotopicmaps in human supplementary motor area. PLOS BIOLOGY,17, e3000026.

Radua, J., Del Pozo, N. O., Gomez, J., Guillen-Grima, F., &Ortuño, F. (2014). Meta-analysis of functional neuroimagingstudies indicates that an increase of cognitive difficultyduring executive tasks engages brain regions associatedwith time perception. Neuropsychologia, 58, 14–22.

Rammsayer, T. H. (1999). Neuropharmacological evidence fordifferent timing mechanisms in humans. The QuarterlyJournal of Experimental Psychology: Comparative andPhysiological Psychology, 52, 273–286.

Rao, S. M., Harrington, D. L., Haaland, K. Y., Bobholz, J. A.,Cox, R. W., & Binder, J. R. (1997). Distributed neural systemsunderlying the timing of movements. Journal ofNeuroscience, 17, 5528–5535.

Rao, S. M., Mayer, A. R., & Harrington, D. L. (2001). Theevolution of brain activation during temporal processing.Nature Neuroscience, 4, 317–q323.

Rekkas, P. V., Westerveld, M., Skudlarski, P., Zumer, J., Pugh, K.,Spencer, D. D., et al. (2005). Neural correlates of temporal-order judgments versus those of spatial-location:Deactivation of hippocampus may facilitate spatialperformance. Brain and Cognition, 59, 103–113.

Riecker, A., Kassubek, J., Gröschel, K., Grodd, W., &Ackermann, H. (2006). The cerebral control of speechtempo: Opposite relationship between speaking rate andBOLD signal changes at striatal and cerebellar structures.Neuroimage, 29, 46–53.

Riecker, A., Wildgruber, D., Mathiak, K., Grodd, W., &Ackermann, H. (2003). Parametric analysis of rate-dependenthemodynamic response functions of cortical and subcorticalbrain structures during auditorily cued finger tapping: A fMRIstudy. Neuroimage, 18, 731–739.

Rosenthal, R. (1979). The file drawer problem and tolerance fornull results. Psychological Bulletin, 86, 638–641.

Rovelli, C. (2018). The order of time. Penguin.Rubia, K., Overmeyer, S., Taylor, E., Brammer, M., Williams, S.,

Simmons, A., et al. (1998). Prefrontal involvement in"temporal bridging" and timing movement.Neuropsychologia, 36, 1283–1293.

Sakai, K., Hikosaka, O., Miyauchi, S., Takino, R., Tamada, T.,Iwata, N. K., et al. (1999). Neural representation of a rhythmdepends on its interval ratio. Journal of Neuroscience, 19,10074–10081.

Schubotz, R. I., Friederici, A. D., & von Cramon, D. Y.(2000). Time perception and motor timing: A commoncortical and subcortical basis revealed by fMRI.Neuroimage, 11, 1–12.

Schubotz, R. I., & von Cramon, D. Y. (2001). Interval andordinal properties of sequences are associated with distinctpremotor areas. Cerebral Cortex, 11, 210–222.

Schwartze, M., Rothermich, K., & Kotz, S. A. (2012). Functionaldissociation of pre-SMA and SMA-proper in temporalprocessing. Neuroimage, 60, 290–298.

Seeley, W. W., Menon, V., Schatzberg, A. F., Keller, J.,Glover, G. H., Kenna, H., et al. (2007). Dissociable intrinsicconnectivity networks for salience processing and executivecontrol. Journal of Neuroscience, 27, 2349–2356.

Shergill, S. S., Tracy, D. K., Seal, M., Rubia, K., & McGuire, P.(2006). Timing of covert articulation: An fMRI study.Neuropsychologia, 44, 2573–2577.

Shih, L. Y., Kuo, W. J., Yeh, T. C., Tzeng, O. J., & Hsieh, J. C.(2009). Common neural mechanisms for explicit timing inthe subsecond range. NeuroReport, 20, 897–901.

Shih, L. Y., Yeh, T. C., Kuo, W. J., Tzeng, O. J., & Hsieh, J. C.(2010). Effect of temporal difficulty on cerebrocerebellarinteraction during visual duration discrimination.Behavioural Brain Research, 207, 155–160.

Singer, T., Critchley, H. D., & Preuschoff, K. (2009). Acommon role of insula in feelings, empathy anduncertainty. Trends in Cognitive Sciences, 13, 334–340.

Smith, A., Taylor, E., Lidzba, K., & Rubia, K. (2003). A righthemispheric frontocerebellar network for timediscrimination of several hundreds of milliseconds.Neuroimage, 20, 344–350.

Smith, A. B., Giampietro, V., Brammer, M., Halari, R., Simmons,A., & Rubia, K. (2011). Functional development of fronto-striato-parietal networks associated with time perception.Frontiers in Human Neuroscience, 5, 136.

Smythies, J., Edelstein, L., & Ramachandran, V. (2012).Hypotheses relating to the function of the claustrum.Frontiers in Integrative Neuroscience, 6, 53.

Smythies, J., Edelstein, L., & Ramachandran, V. (2014).Hypotheses relating to the function of the claustrum II:Does the claustrum use frequency codes? Frontiers inIntegrative Neuroscience, 8, 7.

Stevens, M. C., Kiehl, K. A., Pearlson, G., & Calhoun, V. D.(2007). Functional neural circuits for mental timekeeping.Human Brain Mapping, 28, 394–408.

Taatgen, N. A., van Rijn, H., & Anderson, J. (2007). Anintegrated theory of prospective time interval estimation:The role of cognition, attention, and learning. PsychologicalReview, 114, 577–598.

Taniwaki, T., Okayama, A., Yoshiura, T., Togao, O.,Nakamura, Y., Yamasaki, T., et al. (2006). Functionalnetwork of the basal ganglia and cerebellar motor loops invivo: Different activation patterns between self-initiatedand externally triggered movements. Neuroimage, 31,745–753.

Teki, S. (2016). A citation-based analysis and review ofsignificant papers on timing and time perception. Frontiersin Neuroscience, 10, 330.

Teki, S., & Griffiths, T. D. (2016). Brain bases of workingmemory for time intervals in rhythmic sequences. Frontiersin Neuroscience, 10, 239.

Tipples, J., Brattan, V., & Johnston, P. (2013). Neural bases forindividual differences in the subjective experience of shortdurations (less than 2 seconds). PLOS ONE, 8, e54669.

Tomasi, D., Wang, G. J., Studentsova, Y., & Volkow, N. D.(2015). Dissecting neural responses to temporal prediction,attention, and memory: Effects of reward learning andinteroception on time perception. Cerebral Cortex, 25,3856–3867.

Torgerson, C. M., & Van Horn, J. D. (2014). A case study inconnectomics: The history, mapping, and connectivity of theclaustrum. Frontiers in Neuroinformatics, 8, 83.

Tracy, J. I., Faro, S. H., Mohamed, F. B., Pinsk, M., & Pinus, A.(2000). Functional localization of a “Time Keeper” functionseparate from attentional resources and task strategy.Neuroimage, 11, 228–242.

Tregellas, J. R., Davalos, D. B., & Rojas, D. C. (2006). Effect oftask difficulty on the functional anatomy of temporalprocessing. Neuroimage, 32, 307–315.

Tsukamoto, T., Kotani, Y., Ohgami, Y., Omura, K., Inoue, Y.,& Aihara, Y. (2006). Activation of insular cortex andsubcortical regions related to feedback stimuli in a timeestimation task: An fMRI study. Neuroscience Letters,399, 39–44.

30 Journal of Cognitive Neuroscience Volume X, Number Y

Page 32: AperTO - Archivio Istituzionale Open Access dell ...

Turkeltaub, P. E., Eickhoff, S. B., Laird, A. R., Fox, M., Wiener,M., & Fox, P. (2012). Minimizing within-experiment andwithin-group effects in activation likelihood estimation meta-analyses. Human Brain Mapping, 33, 1–13.

Tzourio, N., Massioui, F. E., Crivello, F., Joliot, M., Renault, B., &Mazoyer, B. (1997). Functional anatomy of human auditoryattention studied with PET. Neuroimage, 5, 63–77.

Unger, R. M., & Smolin, L. (2015). The singular universe andthe reality of time. Cambridge University Press.

Ustun, S., Kale, E. H., & Cicek, M. (2017). Neural networks fortime perception and working memory. Frontiers HumanNeuroscience, 11, 83.

van Wassenhove, V., Wittmann, M., Craig, A. D., & Paulus, M. P.(2011). Psychological and neural mechanisms of subjectivetime dilation. Frontiers Neuroscience, 5, 56.

Vicario, C. M., Martino, D., & Koch, G. (2013). Temporalaccuracy and variability in the left and right posterior parietalcortex. Neuroscience, 245, 121–128.

Westheimer, G. (1999). Discrimination of short time intervalsby the human observer. Experimental Brain Reserach,129, 121–126.

Wiener, M., Lee, Y. S., Lohoff, F. W., & Coslett, H. B. (2014).Individual differences in the morphometry and activation of

time perception networks are influenced by dopaminegenotype. Neuroimage, 89, 10–22.

Wiener, M., Turkeltaub, P., & Coslett, H. B. (2010). The image oftime: A voxel-wise meta-analysis. Neuroimage, 49, 1728–1740.

Wittmann, M. (2013). The inner sense of time: How the braincreates a representation of duration. Nature ReviewsNeuroscience, 14, 217–223.

Wittmann, M., Simmons, A. N., Aron, J. L., & Paulus, M. P. (2010).Accumulation of neural activity in the posterior insula encodesthe passage of time. Neuropsychologia, 48, 3110–3120.

Wittmann, M., Simmons, A. N., Flagan, T., Lane, S. D.,Wackermann, J., & Paulus, M. P. (2011). Neural substrates oftime perception and impulsivity. Brain Research, 1406, 43–58.

Wittmann, M., & van Wassenhove, V. (2009). The experience oftime: Neural mechanisms and the interplay of emotion,cognition and embodiment. Philosophical Transantactionsof the Royal Society of London: Series B: BiologicalSciences, 364, 1809–1813.

Wittmann, M., van Wassenhove, V., Craig, A. D., & Paulus, M. P.(2010). The neural substrates of subjective time dilation.Frontiers in Human Neuroscience, 4, 2.

Zakay, D., & Block, R. A. (1997). Temporal cognition. CurrentDirections in Psychological Science, 6, 12–16.

Nani et al. 31